Federal Communications Commission DA 15-1487 1 Before the Federal Communications Commission Washington, D.C. 20554 In the Matter of Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993 Annual Report and Analysis of Competitive Market Conditions With Respect to Mobile Wireless, Including Commercial Mobile Services ) ) ) ) ) ) ) ) WT Docket No. 15-125 EIGHTEENTH REPORT Adopted: December 23, 2015 Released: December 23, 2015 By the Chief, Wireless Telecommunications Bureau: TABLE OF CONTENTS Heading Paragraph # I. INTRODUCTION .................................................................................................................................. 1 II. COMPETITIVE DYNAMICS WITHIN THE MOBILE WIRELESS INDUSTRY ............................. 8 A. Service Providers .............................................................................................................................. 9 1. Facilities-Based Service Providers ............................................................................................ 9 2. Resellers/Mobile Virtual Network Operators and Other Service Providers ............................ 11 B. Connections, Net Additions, and Churn ......................................................................................... 14 1. Subscribers, Total Connections, and Net Additions ................................................................ 14 2. Subscribers and Net Additions ................................................................................................. 17 3. Churn ....................................................................................................................................... 20 C. Market Shares and Concentration .................................................................................................. 21 D. Financial Indicators ........................................................................................................................ 26 1. Revenue and Average Revenue Per Unit ................................................................................. 26 2. Average Revenue Per Unit by Service Provider ...................................................................... 28 3. Wireless Telephone Services Consumer Price Index .............................................................. 29 4. Profitability Metrics ................................................................................................................. 31 III. OVERALL MOBILE WIRELESS INDUSTRY METRICS ................................................................ 33 A. Network Coverage .......................................................................................................................... 34 1. Overall Mobile Wireless Network Coverage ........................................................................... 36 2. LTE Mobile Broadband Coverage ........................................................................................... 38 3. Rural/Non-Rural Comparisons ................................................................................................ 39 (i) Mobile Wireless Network Coverage ........................................................................... 40 (ii) Mobile LTE Network Coverage ................................................................................. 41 4. Mobile Wireless and LTE Broadband Coverage by Income Levels ........................................ 42 B. Connections and Subscribers .......................................................................................................... 43 1. Connections and Subscribers by Geography ........................................................................... 43 2. Connections and Subscribers by Demographics ...................................................................... 44 Federal Communications Commission DA 15-1487 2 IV. INPUT MARKETS ............................................................................................................................... 46 A. Spectrum ......................................................................................................................................... 47 1. Importance of Spectrum for the Provision of Mobile Wireless Services ................................. 47 2. Facilitating Access to Spectrum ............................................................................................... 49 a. Auctions ............................................................................................................................ 50 b. Secondary Markets ............................................................................................................ 53 c. Additional Spectrum Initiatives ......................................................................................... 58 3. Analysis of Spectrum Holdings ............................................................................................... 60 B. Non-Spectrum Input Segments....................................................................................................... 63 1. Wireless Infrastructure ............................................................................................................. 63 2. Backhaul .................................................................................................................................. 69 3. Recent FCC Initiatives ............................................................................................................. 71 V. PRICING LEVELS AND TRENDS..................................................................................................... 73 A. Postpaid Service ............................................................................................................................. 74 B. Prepaid Service ............................................................................................................................... 94 C. Price Indicators for Mobile Wireless Services ............................................................................. 104 VI. NON-PRICE RIVALRY .................................................................................................................... 105 A. Investment .................................................................................................................................... 106 B. Network Coverage and Technology Upgrades ............................................................................. 109 1. Current Coverage by Service Provider .................................................................................. 110 2. Service Provider Network Deployments................................................................................ 115 3. FCC Initiatives Related to Coverage, Technology, and Roaming ......................................... 121 C. Quality of Service ......................................................................................................................... 125 1. Ookla ...................................................................................................................................... 129 2. FCC ........................................................................................................................................ 130 3. RootMetrics ........................................................................................................................... 131 4. CalSPEED .............................................................................................................................. 132 5. Latency ................................................................................................................................... 133 D. Differentiation in Mobile Wireless Devices and Advertising/Marketing..................................... 136 1. Differentiation in Mobile Wireless Devices .......................................................................... 136 2. Advertising and Marketing .................................................................................................... 141 VII. CONSUMERS AND TRENDS IN THE MOBILE WIRELESS ECOSYSTEM ............................. 143 A. Handsets and Downstream Mobile Applications ......................................................................... 144 B. Usage ............................................................................................................................................ 148 C. Consumer Access to Information ................................................................................................. 151 D. Intermodal Developments ............................................................................................................ 155 VIII.CONCLUSION ................................................................................................................................. 157 IX. PROCEDURAL MATTERS ............................................................................................................ 158 APPENDIX I: MOBILE WIRELESS 3G OR BETTER COVERAGE BY NUMBER OF PROVIDERS APPENDIX II: COMPETITIVE DYNAMICS WITHIN THE INDUSTRY APPENDIX III: OVERALL WIRELESS INDUSTRY METRICS APPENDIX IV: INPUT MARKET APPENDIX V: PRICING APPENDIX VI: NON-PRICE RIVALRY APPENDIX VII: CONSUMERS AND TRENDS IN THE MOBILE WIRELESS ECOSYSTEM I. INTRODUCTION 1. Mobile wireless services are an essential part of Americans’ daily lives. Mobile handsets are no longer used only for voice communication, email, social networking, and web browsing, but are increasingly used as hubs for entertainment, mobile commerce, home automation, and to connect other personal devices such as smart watches, fitness trackers, and health monitors. Further, mobile wireless serves a critical role in public Federal Communications Commission DA 15-1487 3 safety, enabling users to summon lifesaving help, receive timely alerts, and access pertinent information. These developments have made mobile wireless one of the most important sectors in the national economy. Preserving and promoting competition in mobile wireless services is central to the Commission’s mission and is critical for driving innovation and investment to the ultimate benefit of the American consumer. 2. In this Eighteenth Mobile Wireless Competition Report (“Report”), the Federal Communications Commission (“Commission” or “FCC”) fulfills its obligation, pursuant to section 332(c)(1)(C) of the Communications Act (“Act”), to report annually to Congress on the state of competition in mobile services. Following on from the Seventeenth Report, released in December 2014, which provided an analysis of market conditions and developments during 2013 and the first half of 2014,1 this Report presents data and analysis covering the remainder of 2014 and the first half of 2015, to the extent data are available.2 The analysis focuses on “competitive market conditions with respect to commercial mobile services,” as required by the Act.3 3. In its presentation of a multitude of industry data on various aspects of mobile wireless competition,4 this Report follows the model adopted by the Seventeenth Report, which is data-centric, combining a concise analysis with a substantial use of Tables and Charts in accessible data formats.5 As with previous Reports, including the Seventeenth Report, the analysis in this Report is based on a consumer-oriented view of mobile services. This Report therefore follows our practice of undertaking an analysis of all mobile wireless services, including voice, messaging, and broadband.6 Consumers view various mobile voice, messaging, and data services as interchangeable with one another, and as a result, it is important to consider potential substitutes when analyzing the competitive landscape for these services, and to evaluate the mobile wireless industry as a whole, rather than just focusing on the provision of services classified as commercial mobile radio services (“CMRS”).7 Thus, this Report analyzes competition across the entire mobile wireless marketplace, including key market segments such as spectrum and infrastructure. 1 See Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993; Annual Report and Analysis of Competitive Market Conditions With Respect to Mobile Wireless, Including Commercial Mobile Services, WT Docket No. 13-135, Seventeenth Report, 29 FCC Rcd 15311 (WTB 2014) (“Seventeenth Report”). 2 For instance, some of the data are only published as year-end numbers and are publicly available only in the middle of the following year. Quarterly and annual SEC filings for the public wireless service providers are available soon after the release of their financial releases; however, aggregate industry data for public and non-public service providers tend only to be available after they have been compiled by analysts and trade associations based on their set releases. For example, all CTIA – The Wireless Association (“CTIA”) data are now released based on year-end data available in its annual report published a few months after the close of its industry survey. For these data, we are able to present only annualized numbers with no mid-year updates. 3 47 U.S.C. § 332(c)(1)(C). 4 47 U.S.C. § 332(c)(1)(C). As with previous Reports, this Report does not address the merits of any license transfer applications that are currently pending before the Commission or that may be filed in the future, which will be decided based on the record collected in each proceeding. 5 Following that practice, we are providing the charts and tables in the Report and its Appendices, as well as much of the underlying data, on a dedicated website that we intend to update before the release of the next Report as new data becomes available. See FCC Mobile Wireless Competition Reports, available at https://www.fcc.gov/reports- research/reports?og_group_ref_target_id=1638&field_report_series_tid=1733&shs_term_node_tid_depth=All&=Apply. 6 See, e.g., Seventeenth Report, 29 FCC Rcd at 15348 ¶ 71. 7 See, e.g., Seventeenth Report, 29 FCC Rcd at 15314 ¶ 3; Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993; Annual Report and Analysis of Competitive Market Conditions With Respect to Mobile Wireless, Including Commercial Mobile Services, WT Docket No. 11-186, Sixteenth Report, 28 FCC Rcd 3700, 3734 ¶ 20 (2014) (“Sixteenth Report”). We note that previous Reports have therefore included in their analysis a consideration of mobile broadband service before the reclassification of that service in the 2015 Open Internet Order. See Protecting and Promoting the Open Internet, GN Docket No. 14-28, Report and Order on Remand, Declaratory Ruling, and Order, 30 FCC Federal Communications Commission DA 15-1487 4 4. Congress enacted the requirement in 1993 that the Commission report annually on “competitive market conditions with respect to commercial mobile services.”8 At the same time, it created the statutory classification of “commercial mobile services” to promote the consistent regulation of mobile radio services that are similar in nature,9 and established the promotion of competition as a fundamental goal for CMRS policy formation and regulation.10 In particular, the statute requiring the annual report on CMRS competition states: The Commission shall review competitive market conditions with respect to commercial mobile services and shall include in its annual report an analysis of those conditions. Such analysis shall include an identification of the number of competitors in various commercial mobile services, an analysis of whether or not there is effective competition, an analysis of whether any of such competitors have a dominant share of the market for such services, and a statement of whether additional providers or classes of providers in those services would be likely to enhance competition.11 5. This Report complies with the statutory requirements for analyzing competitive market conditions with respect to commercial mobile services. This Report analyzes competition in the mobile wireless services marketplace, as well as examining competition across the entire mobile wireless ecosystem. We analyze the competitive rivalry between service providers in the mobile wireless marketplace and how that competitive rivalry, innovation, and investment benefit American consumers. Consistent with the Commission’s first seven Reports, and the Fourteenth Report and subsequent Reports, this Eighteenth Report does not reach an overall conclusion or formal finding regarding whether or not the CMRS marketplace was effectively competitive, but rather it provides an analysis and description of the CMRS industry’s competitive metrics and trends.12 Given the complexity of the various inter-related segments and services within the mobile wireless ecosystem, any single conclusion regarding the effectiveness of competition would be incomplete and possibly misleading in light of the complexities we observe.13 This Report instead focuses on presenting the best data available on various aspects of competition throughout the mobile wireless ecosystem and highlights several key trends. 6. First, this Report provides an analysis of the overall competitive dynamics of the industry, describing the various operating entities and their relative positions using indices such as market share, subscribership (totals, additions, and churn), as well as various financial indicators.14 The Report then presents a broad overview of industry trends and developments in the mobile marketplace that have taken place since the Rcd 5601, 5715 ¶ 388 (2015) (“2015 Open Internet Order”). 8 47 U.S.C. § 332(c)(1)(C). As noted in previous Reports, any individual proceeding in which the Commission defines relevant product and geographic markets, such as an application for approval of a license transfer, may lead to narrower or broader market(s) than any used, suggested, or implied in this Report. See, e.g., Seventeenth Report, 29 FCC Rcd at 15360 ¶ 100. 9 Omnibus Budget Reconciliation Act of 1993, Pub. L. No. 103-66, Title VI, § 6002(b), amending the Communications Act of 1934 and codified at 47 U.S.C. § 332(c). 10 47 U.S.C. § 332 (a)(3). 11 47 U.S.C. § 332 (c)(1)(C). 12 This is in contrast to the Eighth Report through the Thirteenth Report, which included a specific finding that there was effective competition in the provision of CMRS service without defining the term “effective competition.” See, e.g., Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993; Annual Report and Analysis of Competitive Market Conditions With Respect to Commercial Mobile Services, WT Docket No. 08-27, Thirteenth Report, 24 FCC Rcd 6185, 6310 ¶ 274 (WTB 2009) (“Thirteenth Report”). 13 We note that there is no definition of “effective competition” widely accepted by economists or competition policy authorities such as the U.S. Department of Justice (“DOJ”). See Seventeenth Report, 29 FCC Rcd at 15315 ¶ 6. 14 Unless stated otherwise, dollar figures stated in this Report have not been adjusted for inflation (i.e., they are nominal dollars). Federal Communications Commission DA 15-1487 5 Seventeenth Report, such as increased network coverage, subscribership growth, and the adoption and deployment of technologies. The Report then turns to an analysis of key inputs necessary for provision of mobile service, such as spectrum and network infrastructure. Spectrum, in particular, is a critical input that wireless service providers need for the provision of mobile wireless services, and this Report examines the distribution of spectrum in the various bands. Next, the Report analyzes recent developments in the ways service providers compete for and attract subscribers through pricing innovations, such as the decreased reliance on traditional handset subsidies and term contracts. The Report then analyzes competitive rivalry in non-price factors, such as the coverage, service quality, and speed of providers’ service offerings. Finally, the Report considers developments in the downstream mobile wireless ecosystem as well as issues such as consumer access to information and intermodal developments. 7. In addition to providing an analysis of market conditions, the Report highlights the Commission’s policies and actions designed to enhance competition. For example, the Commission is making more spectrum available to existing mobile service providers and potential new entrants through competitive bidding, such as in the upcoming Incentive Auction and promoting innovative new approaches, which includes spectrum sharing in the 3.5 GHz Band.15 The Commission’s policies have been guided by the goal of promoting and preserving competition, which in turn has facilitated the ability of consumers to make choices among numerous service providers and leads to lower prices, improved quality, and increased innovation.16 II. COMPETITIVE DYNAMICS WITHIN THE MOBILE WIRELESS INDUSTRY 8. We begin our analysis by providing a discussion of the various competitive dynamics within the industry.17 We note that providers of mobile wireless services typically offer an array of mobile voice and data services, including interconnected mobile voice services, text and multimedia messaging, and mobile broadband Internet access services.18 Facilities-based mobile wireless service providers offer such services primarily using their own network facilities, although coverage areas often are supplemented through roaming agreements, and may operate nationwide, multi-regional, regional, or local networks.19 In this section, we present information and 15 See Broadcast Incentive Auction Scheduled to Begin on March 29, 2016, Procedures for Competitive Bidding in Auction 1000, Including Initial Clearing Target Determination, Qualifying To Bid, and Bidding in Auctions 1001 (Reverse) and 1002 (Forward), AU Docket No. 14-252, GN Docket No. 12-268, WT Docket No. 12-269, MB Docket No. 15-146, Public Notice, 30 FCC Rcd 8975 (2015) (“Auction 1000 Bidding Procedures Public Notice”); Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Report and Order and Second Further Notice of Proposed Rulemaking, 30 FCC Rcd 3959 (2015) (“3.5 GHz Order and 2nd FNPRM”). 16 See Policies Regarding Mobile Spectrum Holdings; Expanding the Economic and Innovation Opportunities of Spectrum Through Incentive Auctions, WT Docket No. 12-269, GN Docket No. 12-268, Report and Order, 29 FCC Rcd 6133, 6144 ¶ 17, 6193 ¶ 143 (2014) (“Mobile Spectrum Holdings Report and Order”), recon. denied, Order on Reconsideration, 30 FCC Rcd 8635 (2015) (“Mobile Spectrum Holdings Reconsideration Order”). Our public interest evaluation necessarily encompasses the “broad aims of the Communications Act,” which include, among other things, a deeply rooted preference for preserving and enhancing competition in relevant markets, accelerating private sector deployment of advanced services, and generally managing the spectrum in the public interest. See, e.g., Applications of AT&T Mobility Spectrum LLC, New Cingular Wireless PCS, LLC, Comcast Corporation, Horizon Wi-Com, LLC, NextWave Wireless, Inc., and San Diego Gas & Electric Company for Consent To Assign and Transfer Licenses, WT Docket No. 12-240, Memorandum Opinion and Order, 27 FCC Rcd 16459, 16472 ¶ 34 (2012) (“AT&T WCS Order”). 17 We discuss in sections V. and VI. below additional aspects in the competitive dynamics of the industry when we discuss elements of price and non-price rivalry. 18 We note that mobile wireless services also include machine-to-machine connections for fleet management systems, smart grid devices, vehicle tracking, home security systems, and other telematics services. Fixed wireless services currently are not included in our analysis of mobile wireless services. 19 Some data and messaging services offered by facilities-based providers rely only on Internet Protocol (“IP”)-based, packet- switched networks, but most mobile voice services continue to connect to the Public Switched Telephone Network (“PSTN”) and rely on North American Numbering Plan (“NANP”) telephone numbers. Federal Communications Commission DA 15-1487 6 data on all mobile wireless services as well as on individual services and segments where appropriate, and where the data are available. A. Service Providers 1. Facilities-Based Service Providers 9. Nationwide Service Providers. As of year-end 2014, there were four facilities-based mobile wireless service providers in the United States that industry observers typically describe as “nationwide”: AT&T,20 Sprint,21 T-Mobile,22 and Verizon Wireless.23 Although none of these four service providers has a network that covers the entire land area or population of the United States, each has a network that covers a significant portion of both and therefore these four service providers will be referred to as “nationwide service providers” throughout this Report.24 10. Multi-Regional, Regional, and Local Service Providers. US Cellular is a multi-regional service provider that has developed wireless networks and customer service operations in portions of 23 states and as of December 31, 2014, provided services to approximately 4.8 million customers.25 US Cellular relies on roaming agreements with nationwide service providers, as well as other smaller service providers, to supply service to its 20 AT&T Mobility began operations in October 2000 as a joint venture between AT&T and BellSouth and, in 2004, acquired AT&T Wireless Services, Inc. Upon AT&T’s acquisition of BellSouth in 2006, AT&T Mobility became a wholly-owned subsidiary. See AT&T Inc., Form 10-K, for the fiscal year ended December 31, 2014, at 2, available at https://www.sec.gov/Archives/edgar/data/732717/000073271715000016/ye14_10k.htm. 21 Sprint Nextel was created by the merger in 2005 of Sprint Corp. and Nextel Communications, Inc. See Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993; Annual Report and Analysis of Competitive Market Conditions With Respect to Commercial Mobile Services, WT Docket No. 05-71, Tenth Report, 20 FCC Rcd 15908, 15931 ¶ 60 (2005). On July 5, 2013, the Commission released an order approving the acquisition of Sprint by SoftBank Corp., and Sprint’s acquisition of 100% of Clearwire’s stock. See Applications of Sprint Nextel Corp. and SoftBank Corp. and Starburst II, Inc. for Consent To Transfer Control of Licenses and Authorizations, IB Docket No. 12-343, Memorandum Opinion and Order, Declaratory Ruling, and Order on Reconsideration, 28 FCC Rcd 9642, 9643-44 ¶¶ 1-4 (2013) (“Softbank-Sprint- Clearwire Order”). 22 T-Mobile traces its roots to May 2001,when Deutsche Telekom AG acquired VoiceStream Wireless and Powertel. In September 2002, they were re-branded with the T-Mobile brand name. See T-Mobile USA Inc., Form 10-K, for the fiscal year that ended December 31, 2002, at 2, available at http://www.sec.gov/Archives/edgar/data/1097609/000089102003000720/v88048ore10vk.htm. Most recently, on March 12, 2013, the application of Deutsche Telekom, T-Mobile, and MetroPCS was approved, which resulted in the creation of T- Mobile USA as a wholly-owned subsidiary of Deutsche Telekom. See Applications of Deutsche Telekom AG, T-Mobile USA, Inc., and MetroPCS Communications, Inc. for Consent To Transfer of Control of Licenses and Authorizations, WT Docket No. 12-301, Memorandum Opinion and Order and Declaratory Ruling, 28 FCC Rcd 2322, 2323-24 ¶¶ 1-2 (WTB, IB 2013) (“T-Mobile-MetroPCS Order”). 23 As of December 31, 2013, Verizon owned a controlling 55% interest in Verizon Wireless and Vodafone owned the remaining 45%. On September 2, 2013, Verizon entered into a stock purchase agreement with Vodafone and Vodafone 4 Limited, pursuant to which Verizon agreed to acquire Vodafone’s indirect 45% interest in Cellco Partnership d/b/a Verizon Wireless for approximately $130 billion. Verizon completed the transaction on February 21, 2014 and acquired 100% ownership of Verizon Wireless. See Verizon Communications, Inc., SEC Form 10-K, for the fiscal year that ended December 31, 2013, at 2, available at https://www.sec.gov/Archives/edgar/data/732712/000119312514073266/d622994d10k.htm. 24 All four nationwide service providers hold spectrum in the continental United States, as well as in Hawaii and Alaska. 25 See United States Cellular Corp., SEC Form 10-K, at 1 (filed Feb. 25, 2015), available at http://www.sec.gov/Archives/edgar/data/821130/000082113015000010/form_10k.htm. USCC is a majority-owned (84%) subsidiary of Telephone and Data Systems, Inc. See id. Federal Communications Commission DA 15-1487 7 customers when they travel to areas not covered by its networks. C Spire and nTelos26 are two other regional service providers with substantial market presence in certain parts of the country. There are also dozens of regional and local facilities-based service providers27 throughout the continental United States, Alaska, and Hawaii that typically provide service in a single geographical area, many of them rural areas.28 As the Commission noted in the Mobile Spectrum Holdings Report and Order, non-nationwide service providers are important sources of competition in rural areas, enhancing competitive choices for consumers in the mobile wireless marketplace, and helping to promote deployment in rural areas.29 2. Resellers/Mobile Virtual Network Operators and Other Service Providers 11. Resellers/MVNOs. Resellers and mobile virtual network operators (“MVNOs”) do not own any network facilities, but instead they purchase mobile wireless services wholesale from facilities-based providers and resell these services to consumers.30 An agreement between an MVNO and a facilities-based service provider may be more likely to occur when the MVNO has better access to some market segments than the host facilities- based service provider, and can potentially target specific market segments such as low-income consumers or consumers with low-usage needs.31 Unlike facilities-based service providers, MVNOs do not engage in non-price rivalry by creating capacity through network investments, network upgrades, or network coverage. In 2014, the largest MVNO, with approximately 26 million subscribers at year-end, was TracFone Wireless (“TracFone”).32 26 nTelos recently entered into agreements with Shenandoah Personal Communications, LLC (“Shenandoah”) and Sprint pursuant to which nTelos would become a wholly-owned subsidiary of Shenandoah, and all of nTelos’s spectrum licenses would be assigned to Sprint. See Application To Assign Licenses from NTELOS Inc. to SprintCom, Inc., ULS File No. 0006917154 (lead application) (filed Aug. 25, 2015). 27 Some regional facilities-based service providers include, but are not limited to, Alaska Communications Systems, Bluegrass Cellular, Carolina West Wireless, Cellcom, Choice Wireless, Nex-Tech Wireless, Pioneer, and Sagebrush Cellular. 28 Verizon Wireless’s LTE in Rural America (“LRA”) program allows Verizon Wireless to offer its customers 4G LTE coverage in the rural areas of its rural partners, and the program allows customers of participating companies to roam on Verizon Wireless’s 4G LTE network throughout the U.S., including Alaska. See “Verizon’s LTE in Rural America (LRA) Program Celebrates Five Years of Delivering Advanced Wireless Services to Rural Customers,” available at http://www.verizonwireless.com/news/article/2015/05/verizons-lte-in-rural-america-lra-program-celebrates-five-years-of- delivering-advanced-wireless-services-to-rural-customers.html; “Verizon: All 21 LTE in Rural America Carrier Partners Have Launched Service,” available at http://www.fiercewireless.com/story/verizon-all-21-lte-rural-america-carrier-partners- have-launched-service/2015-10-15. Sprint has partnered with the Competitive Carriers Association (“CCA”) and Net America to launch the Net America alliance, which should allow Sprint customers to obtain 4G LTE service via roaming in rural areas and allow customers of participating companies to roam on Sprint’s 4G LTE network throughout the U.S. See Net America Alliance website, “SMART Delivers the Future to Rural America,” available at http://www.netamericaalliance.com/smart. 29 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6207 ¶¶ 179-80. 30 According to Verizon Wireless, “MVNOs execute a contract with [the facilities-based provider] to buy wireless service from [the facilities-based provider] to resell under their own brand to customers and perform all marketing, billing, collections and customer service for the customers they activate. MVNOs establish and maintain the relationship with its customers. MVNOs own the relationship with their customers and establish their own calling plans and pricing.” See Verizon Wireless, “Authorized Retailers and MVNOs,” available at http://www.verizonwireless.com/b2c/aboutUs/reseller/authorizedAgentIndex.jsp. 31 See P. Kalmus and L. Wiethaus, “On the Competitive Effects of Mobile Virtual Network Operators,” Telecommunications Policy, Vol. 34, 2010, at 263, 266, 268; A. Banerjee and C. Dippon, “Voluntary Relationships Among Mobile Network Operators and Mobile Virtual Network Operators: An Economic Explanation, Information Economics and Policy,” Vol. 21, 2009, at 72. See also The Yankee Group, Jason Armitage, “Yankee Group’s 2011 Predictions: 4G Fuels the Decade of Disruption,” at 7 (stating, “[I]t’s critical the MVNO does not compete to any meaningful degree with the host.”). 32 See TracFone website, available at http://www.tracfone.com/; http://www.prepaidphonenews.com/2015/02/fouth-quarter- 2014-prepaid-mobile.html. Examples of other MVNO companies include, but are not limited to, Straight Talk, H2O Federal Communications Commission DA 15-1487 8 12. Mobile Satellite Service Providers. Mobile Satellite Services (“MSS”) providers offer satellite- based communications to mobile devices. Traditionally, MSS has involved voice and narrowband data services. MSS services are generally targeted at users who require service in remote areas, in disaster response situations, or other places where terrestrial mobile wireless network access may be limited. Examples of MSS customers include the oil industry, maritime users, public safety agencies, and other government/military operations. 13. Narrowband Data Service Providers. Narrowband data and paging services comprise a specialized market segment of the mobile wireless industry. These services include two-way messaging, as well as machine-to-machine and other telemetry communications, and are consumed primarily by businesses, government users, and other institutions.33 B. Connections, Net Additions, and Churn 1. Subscribers, Total Connections, and Net Additions 14. This Report uses several data sources to estimate the number of mobile wireless subscribers and connections. One source, the Numbering Resource Utilization Forecast (“NRUF”), tracks the quantity of phone numbers that have been assigned to mobile wireless devices.34 As shown in Chart II.B.1 below, in the period since the Seventeenth Report, the total number of connections continued to grow strongly. Based on NRUF, the number of mobile wireless connections in December 2013 were approximately 340 million, and connections grew by approximately five percent during 2014 to reach approximately 357 million by year-end 2014. CTIA also estimated the total number of mobile wireless connections,35 and found that the number of connections grew by approximately six percent during that same period, from approximately 336 million at year-end 2013 to approximately 355 million at year-end 2014.36 Chart II.B.2 presents data on total connections by service segment and shows that in 2014, the postpaid segment accounted for more than 60 percent of the total connections, while the prepaid connections accounted for approximately 20 percent of the total connections. Wholesale connections and connected devices are a small but growing part of total mobile wireless connections.37 Similar patterns were observed through the first half of 2015. Wireless, Ultra Mobile, Net10, LycaMobile, Telcel America, Simple Mobile, Red Pocket, Pure Talk, Ting, iWireless, and Voyager. See Amgoo, “Mobile Industry Insider,” MVNO list, available at http://www.amgoo.com/blog/the-big-mvno-list- 83-mvno-providers-in-usa. 33 There is approximately 7 megahertz of spectrum allocated to narrowband and paging services and there are hundreds of licensees for these services, including private individuals, firms, and local and state governments. 34 When all mobile wireless devices were assigned telephone numbers and subscribers generally carried one mobile device for making voice calls, NRUF was a reasonably accurate measure of subscribership. Currently, however, consumers are more likely to use more than one mobile device that have been assigned telephone numbers – particularly non-voice devices, such as Internet access devices (e.g., wireless modem cards and mobile Wi-Fi hotspots), e-readers, tablets, and telematics systems. In addition, certain mobile broadband service providers do not assign telephone numbers to at least some of the devices on their networks. Therefore, NRUF is becoming less useful in measuring the number of individual subscribers. Instead, it provides a measure of the number of mobile wireless connections or connected devices, although we note that it will become a less accurate measure of connected devices to the extent that more are sold that do not use telephone numbers. 35 CTIA states that “the terms subscriber, subscriptions, and connections are being used interchangeably” in their report and survey. See CTIA Year-End 2014 Wireless Indices Report, at p. 12. 36 See Appendix Table II.B.i for detailed data on total mobile wireless connections. 37 See Appendix Table II.B.ii for detailed data on total mobile wireless connections by service segment. Federal Communications Commission DA 15-1487 9 Chart II.B.1 Total Mobile Wireless Connections: 2008 – 2014 Source: NRUF and CTIA (CTIA Year-End 2014 Wireless Indices Report, Table 6). Chart II.B.2 Quarterly Total Mobile Wireless Connections by Service Segment: 2012 – 1st Half 2015 Source: UBS Investment Research. US Wireless 411 Version 51, Figure 17: US Wireless 411 Version 57, Figure 31. 2008 2009 2010 2011 2012 2013 2014 NRUF 279.6 290.7 301.8 317.3 329.2 339.2 357.1 CTIA 270.3 285.6 296.3 316.0 326.5 335.7 355.4 0 50 100 150 200 250 300 350 400 E st im a te d C o n n ec ti o n s (i n M il li o n s) Year NRUF CTIA 323,056 324,687 326,644 330,729 335,974 337,156 342,566 347,979 352,523 356,809 364,017 371,471 376,213 382,307 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 T o ta l C o n n ec ti o n s (t h o u sa n d s) Quarter, YearPostpaid Prepaid Wholesale Connected devices Federal Communications Commission DA 15-1487 10 15. Table II.B.1 presents data on total mobile wireless connections of the larger service providers operating in the United States. This Table shows that as of year-end 2014, and in mid-2015, Verizon Wireless and AT&T together accounted for approximately two-thirds of the estimated connections, while T-Mobile and Sprint together accounted for slightly less than one-third. As of year-end 2014, T-Mobile had approximately 55 million connections, compared to approximately 56 million for Sprint. By mid-2015, T-Mobile had seen an increase in its number of connections to approximately 59 million, and Sprint had also seen an increase in its connections to approximately 58 million. By mid-2015, regional service providers accounted for well under two percent of total connections.38 Table II.B.1 Estimated Total Connections for Publicly Traded Facilities–Based Mobile Wireless Service Providers (in thousands): 2012 – 1st Half 2015 Nationwide Service Providers EOY 2012 EOY 2013 EOY 2014 June 30, 2015 Verizon Wireless 116,570 125,535 134,612 136,641 AT&T 106,965 110,276 120,620 123,932 T-Mobile 30,299 46,684 55,018 58,908 Sprint 55,626 54,622 55,929 57,668 Nationwide Service Provider Total 309,460 337,117 366,179 377,149 Regional Service Providers EOY 2012 EOY 2013 EOY 2014 June 30, 2015 US Cellular 5,798 4,774 4,760 4,779 Metro PCS 8,887 * * * Leap Wireless 5,297 4,551 * * NTELOS 440 465 449 379 Cincinnati Bell 398 340 82 * Regional Service Provider Total 20,820 10,130 5,291 5,158 Total Estimated Connections 330,279 347,247 371,470 382,307 Source: UBS US Wireless 411 Report. Version 51 2014 Q1, Table 21. UBS US Wireless 411 Version 57, Figure 41. Total estimated connections figure includes data only for the service providers reported in this table. 16. All four nationwide service providers provide service directly to consumers and businesses and also provide machine-to-machine (“M2M”) services.39 Later in this Report, detailed data and analysis are provided on retail voice and broadband services; however, there are limited statistics on M2M communications. For 2014, Berg Insight, for example, reports that AT&T led the U.S. market with approximately 16 million M2M 38 We note that C Spire, the largest privately held service provider in the U.S., whose total number of connections are not reflected in this Table, states that it has nearly 1 million subscribers, with its primary service area in the Southeastern U.S. See “C Spire, Who We Are,” available at http://www.cspire.com/company_info/about/more_info.jsp. 39 M2M is a subset of the larger Internet of Things (“IoT”), and aside from differing definitions, researchers may be including or excluding connections that are not specifically defined by the industry as M2M. These variations make it difficult to compare data from multiple reported sources. The IoT is seen by some commentators as the next major opportunity for providing advanced connections among devices, and many industries such as healthcare are beginning to transform to use M2M networks to connect their numerous smart devices and machines. Federal Communications Commission DA 15-1487 11 subscribers, ahead of Verizon Wireless at approximately 9 million.40 SNL Kagan reports that in the second quarter of 2015, AT&T had a connected device base of 23.4 million, Verizon Wireless had 16.3 million, Sprint had 8 million, and T-Mobile had 4.5 million.41 Many research firms forecast that the overall trends for M2M will become more significant as new and existing network service providers continue to deliver connectivity between devices, sensors, monitors, etc., and their networks. 2. Subscribers and Net Additions 17. As shown in Chart II.B.3 below, net additions for 2014 totaled approximately 18 million based on NRUF data, and approximately 20 million based on CTIA data. In addition, we include mobile voice and internet subscriber data as reported by service providers on Form 477, which generally show a lower number of subscriber additions than NRUF. For 2014, Form 477 shows that net subscriber additions totaled approximately 12 million. Source: NRUF and CTIA (CTIA Wireless Industry Indices Year-End 2014, Table 6), Form 477. 40 See PR Newswire, “The Global Wireless M2M Market – 6th Edition,” available at http://www.prnewswire.com/news- releases/the-global-wireless-m2m-market--6th-edition-300022274.html. Further, Chetan Sharma Consulting reports that total M2M subscribers for the U.S. were 22.7 million at year-end 2014, which increased by 1.1 million (or approximately 5%) by 2Q 2015 to 23.8 million. They noted that one of the major growth areas is the connected car segment in which major manufacturers have started including wireless connectivity in their offerings. See http://www.chetansharma.com/research.htm. 41 SNL Kagan defines connected devices to include tablets/eReaders with 3G/4G, cars with 3G/4G, wearables, eHealth, etc. See SNL Kagan: “Smart pipes: Mobile projections through 2025” (Oct. 12, 2015) (complete list of connections included in its examples of connected devices). 11.1 11.1 15.5 11.9 10.0 17.9 15.3 10.7 19.7 10.5 9.2 19.8 13.0 10.8 12.2 7.6 5.9 11.8 0.0 5.0 10.0 15.0 20.0 25.0 2009 2010 2011 2012 2013 2014 N et A d d it io n s (i n M il li o n s) Year Chart II.B.3 Total Mobile Wireless Connections Annual Net Additions (2008-2014) NRUF CTIA Form 477 Federal Communications Commission DA 15-1487 12 18. Chart II.B.4 below shows that the net number of connected device additions was consistently higher than prepaid additions during through the first half of 2014.42 Postpaid net additions showed significant growth during 2014, and while there was a slight dip in the first quarter of 2015, postpaid net additions grew in the second quarter of 2015. Chart II.B.4 also shows that prepaid additions have continued to decline as a percentage of total quarterly net additions. Chart II.B.4 Quarterly Net Additions by Service Segment: 2013 – 1st Half 2015 Source: UBS Investment Research. UBS US Wireless 411 Version 57, Figure 31. 19. Chart II.B.5 below shows net subscriber additions by the four nationwide service providers from 2010 through the first half of 2105. While Chart II.B.5 shows that AT&T and Verizon Wireless continue to show strong growth, of particular note is T-Mobile, which nearly doubled its net additions between 2013 and 2014, and this trend in net additions has continued strongly into 2015. In addition, we note that while Sprint’s net subscriber additions were negative in 2013, its net additions were positive in 2014. Further, through the first half of 2015, its net subscriber additions jumped sharply. 42 See Appendix Table II.B.iii for detailed data on quarterly net additions by service segment. 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 Total 2,181 788 3,229 5,389 4,378 4,418 7,214 7,465 4,748 6,136 Prepaid 1,278 -3,208 1,102 1,098 421 -966 686 712 187 -21 Wholesale 3,431 598 436 802 -945 975 1,497 938 1,088 1,339 Connected Devices 1,344 3,134 1,264 894 1,537 1,434 1,967 2,028 1,471 2,009 Postpaid -3,872 264 427 2,595 3,364 2,976 3,064 3,787 2,003 2,809 -4,000 -2,000 0 2,000 4,000 6,000 8,000 N et A d d s (t h o u sa n d s) Federal Communications Commission DA 15-1487 13 Chart II.B.5 Annual Net Additions by Service Provider: (2010 – 1st Half 2015) Source: UBS Investment Research. UBS US Wireless 411 Version 51, Figure 14: UBS US Wireless 411 Version 57, Figure 50. 3. Churn 20. Churn measures the number of connections that are disconnected from mobile wireless service during a given period time period, and is usually expressed as a percentage.43 Service providers publish their monthly churn rate information as part of their quarterly filings with the SEC. A service provider’s churn rate depends on many factors, including the distribution of its customers between postpaid and prepaid service plans, customer satisfaction with their service provider, service provider switching costs, and competition. The average industry monthly churn rates from the first quarter of 2012 to the second quarter of 2015 have ranged from 1.44 percent to 1.85 percent.44 Churn rates of the nationwide facilities-based service providers, as shown in Chart II.B.6 below, ranged from approximately 1 percent for Verizon Wireless and AT&T to approximately 2 percent for T-Mobile and Sprint for the second quarter of 2015. 43 Churn is calculated by dividing the aggregate number of wireless subscriber connections who canceled service during a period by the total number of wireless subscriber connections at the beginning of that period. The churn rate for the period is equal to the weighted average of the churn rate for each month of that period, e.g., the three months in a quarter or the twelve months for an annual churn rate. Thus, a monthly churn rate of 1% averaged over the three-month reporting period would also be reported as 1%. For example, if a service provider has an average monthly churn rate of 2% in each month of a year, the service provider would lose approximately 24% of its subscriber connections over the course of the year. 44 See UBS Investment Research. US Wireless 411 Version 57, Figure 48. 2010 2011 2012 2013 2014 1stHalf2015 Verizon 5,627 6,288 8,265 8,868 9,235 2,077 AT&T 8,853 7,699 3,764 2,721 5,608 3,312 Sprint 1,777 5,111 605 -2,500 575 1,739 T-Mobile -56 -549 204 4,300 8,334 3,890 -4,000 -2,000 0 2,000 4,000 6,000 8,000 10,000 N et S u b sc ri b er A d d it io n s (T h o u sa n d s) Federal Communications Commission DA 15-1487 14 Source: UBS Investment Research. US Wireless 411 Version 49, Table 16. US Wireless 411 Version 51, Figure 28. US Wireless 411 Version 57, Figure 22. C. Market Shares and Concentration 21. Revenues and connections or subscribers, presented in Table II.C.1 below, are key metrics that are used to measure the size of a company. In turn, the size of a company relative to the total size of the industry determines market share, which is generally calculated as the percentage of an industry or market’s total revenues earned (or number of customers served) by a particular company over a specified time period. In general, changes in market share may provide a signal of the relative competitiveness of a company's products or services. Nationwide (and regional) service provider market shares by service revenues are shown in Table II.C.2 below. The four nationwide service providers accounted for approximately 98 percent of the nation’s mobile wireless service revenue in 2014, up from approximately 91 percent in 2012, and the service revenues of AT&T and Verizon Wireless together accounted for approximately 71 percent of total service revenue in 2014. Of the four nationwide facilities-based service providers, AT&T and Verizon Wireless continued to maintain the largest market shares throughout 2014, as shown in Table II.C.2. Sprint stayed relatively flat, and T-Mobile had the largest quarterly increases in market share to end 2014, as measured by revenue, narrowing the gap against Sprint.45 The same pattern continued in the first half of 2015, with AT&T and Verizon Wireless continuing to account for approximately 71 percent of total service revenues. While T-Mobile continues to narrow the gap against Sprint, as of mid-2015, Sprint remained the third largest service provider in the mobile wireless marketplace in terms of service revenues. 45 See UBS Investment Research. US Wireless 411 Version 55 4Q14, Figure 45. 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 AT&T 1.5% 1.2% 1.3% 1.4% 1.4% 1.4% 1.3% 1.4% 1.4% 1.5% 1.4% 1.6% 1.4% 1.30% Verizon Wireless 1.2% 1.1% 1.2% 1.2% 1.3% 1.2% 1.3% 1.2% 1.2% 1.2% 1.3% 1.4% 1.3% 1.20% Sprint 2.2% 2.0% 2.1% 2.3% 2.5% 3.2% 2.3% 2.2% 3.2% 2.4% 2.3% 2.4% 2.1% 2.25% T-Mobile 3.3% 3.2% 3.4% 3.7% 3.3% 2.6% 2.5% 2.5% 2.6% 2.1% 2.2% 2.5% 2.0% 2.00% 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% C h u rn R a te Chart II.B.6 Quarterly Churn Rate for Nationwide Mobile Wireless Service Providers 1st Quarter 2012 -- 2nd Quarter 2015 Federal Communications Commission DA 15-1487 15 Table II.C.1 Service Revenues for Mobile Wireless Service Providers ($ millions), 2007 – 1st Half 2015 National 2007 2008 2009 2010 2011 2012 2013 2014 1st Half 2015 Verizon 38,016 49,717 52,046 55,629 59,157 63,733 69,033 72,630 35,603 AT&T 38,678 44,249 48,563 53,510 56,726 59,186 61,552 61,032 29,927 Sprint 32,106 28,435 25,832 25,894 27,390 29,086 29,263 27,959 13,178 T-Mobile 16,891 19,242 18,926 18,689 18,481 17,213 20,535 22,375 11,963 Regional 2007 2008 2009 2010 2011 2012 2013 2014 1st Half 2015 US Cellular 3,679 3,940 3,926 3,913 4,054 4,099 3,595 3,398 1,652 NTELOS 357 392 400 383 395 424 467 445 199 Cincinnati Bell 267 291 284 269 252 225 185 126 Leap Wireless 1,396 1,709 2,171 2,413 2,829 2,947 2,631 Metro PCS 1,919 2,437 3,130 3,690 4,428 4,540 Centennial 484 524 408 CentennialPCS 294 320 236 Rural Cellular 608 327 Alltel 7,984 Dobson 1,030 SunCom 649 Source: UBS Investment Research. UBS US Wireless 411 Report. Version 51, Table 31. UBS US Wireless 411 Report Version 57, Figure 51. Table II.C.2 Market Shares for Mobile Wireless Service Providers, Based on Service Revenues, 2012 – 1st Half 2015 Nationwide Service Providers 2012 2013 2014 1st Half 2015 Verizon Wireless 34.4% 36.5% 38.7% 38.1% AT&T 32.0% 32.5% 32.5% 32.6% Sprint 15.7% 15.5% 14.9% 14.1% T-Mobile 9.3% 10.9% 11.9% 13.2% Total National Service Provider Market Share 91.5% 95.3% 97.9% 98.0% Regional Service Providers 2012 2013 2014 1st Half 2015 US Cellular 2.2% 1.9% 1.8% 1.8% Metro PCS 2.5% * * Leap Wireless 1.6% 1.4% * NTELOS 0.2% 0.2% 0.2% 0.2% Cincinnati Bell 0.1% 0.1% 0.1% * Other 1.9% 1.0% * * Total Regional Service Provider Market Share 8.5% 4.7% 2.1% 2.0% Source: UBS US Wireless 411 Report. Version 51 2014 Q1, Table 31, p.19. UBS US Wireless 411 Report. Version 57, Figure 51. See also Seventeenth Report, Table II.C.2 for pre-2014 data. Federal Communications Commission DA 15-1487 16 22. Tables II.C.1 and II.C.2 above also show that over time, regional service providers have accounted for increasingly less of overall industry revenues. Based on service revenues, the market share for regional service providers fell from around 8.5 percent in 2012 to around 2 percent by year-end 2014.46 In addition, Chart II.C.1 below measures market shares using the number of subscribers/connnections as a percentage of overall industry subscribers/connections. It shows that the Top 4 providers have increased their share of overall industry subscribers/connections from around 66 percent in 2003 to around 98.5 percent by year- end 2014,47 meaning that the share of regional and local providers has declined from around 34 percent to around 1.5 percent during the same time period. Chart II.C.1 U.S. Mobile Wireless Connections: 2003 – 2014 Source: 9th Report, Appendix p. A-8, Table 4; 11th Report, p. 93, Appendix Table 4; 12th Report, p. 132, Appendix Table 4; 14th Report, p. 223, Appendix Table C-4; 15th Report, p. 34, Table 3; 16th Report, p. 55, Table 13; 17th Report, p. 10, Table II.B.1; UBS US Wireless 411 Version 57, Figure 41. 23. Market concentration can be measured by the number of competitors in the marketplace or by the sum of the share of subscribers and sales/revenues attributable to each competitor. High market concentration levels in a given market may raise some concern that the market is not competitive. However, an analysis of other factors, such as prices, non-price rivalry, and entry conditions, may find that a market with high concentration levels is competitive. The Herfindahl-Hirschman Index (“HHI”), which is employed by the Commission to measure market concentration, is a widely-accepted measure of concentration in competition analysis. The HHI is calculated by summing the squared market shares of all firms in any given market.48 In this 46 We note that these estimates are based on UBS US Wireless 411 Reports, which do not provide a break out number for privately held regional service provider, C Spire. 47 For purposes of Chart II.C.1, a Top 4 service provider is defined as a service provider that is one of the leading four providers in any given year as measured by its subscribers/connections over total industry subscribers/connections. 48 Following widespread industry practices, the Commission generally attributes the subscribers of MVNOs to their host facilities-based service providers, including when it calculates market concentration metrics. 0% 20% 40% 60% 80% 100% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 P er ce n ta g e o f T o ta l U .S . C o n n ec ti o n s Top 4 Providers All Other Providers Federal Communications Commission DA 15-1487 17 Report, we calculate HHIs by Economic Area (“EA”) to maintain continuity with past Reports and to ensure that we do not compromise the confidential information found in the NRUF data.49 24. As shown in Chart II.C.1, at the end of 2014, the weighted average HHI (weighted by population across the 172 EAs in the United States) for mobile wireless services was 3,138, an increase from 3,027 at year- end 2013.50 As in previous years, the most recent increases in the weighted average HHI reflect continued industry consolidation, such as the acquisition by AT&T of Leap Wireless in 2014.51 At the end of 2014, the HHI in individual EAs ranged from a low of 2,219 in EA 63 (Milwaukee-Racine, WI) to a high of 7,312 in EA 142 (Scottsbluff, NE-WY).52 25. Chart II.C.2 below shows the relationship between the HHI by EA and EA population densities, indicating that HHI values, or market concentration, tend to decline as the population density increases. The most concentrated EAs tend to be more rural, while major metropolitan areas lie in the least concentrated EAs. This likely reflects greater demand and greater cost efficiencies (per-user mobile wireless network deployment costs tend to decrease with increases in the population density) in more densely-populated areas.53 49 NRUF subscriber data indicate the number of assigned phone numbers that a wireless service provider has in a particular wireline rate center (there are approximately 18,000 rate centers in the country). Rate centers are geographic areas used by local exchange carriers for a variety of reasons, including the determination of toll rates. See Harry Newton, Newton’s Telecom Dictionary: 19th Expanded & Updated Edition 660 (July 2003). All mobile wireless service providers must report to the Commission the quantity of their phone numbers that have been assigned to end users, thereby permitting the Commission to calculate the total number of mobile wireless subscribers. For purposes of geographical analysis, the rate center data can be associated with a geographic point, and all of those points that fall within a county boundary can be aggregated together and associated with much larger geographic areas based on counties. We note that the aggregation to larger geographic areas reduces the level of inaccuracy inherent in combining non-coterminous areas, such as rate center areas and counties. As discussed in this Report, “markets” are independent of the relevant market determined in the context of secondary market transactions review. In prior transactions, the Commission has found that the relevant geographic markets for certain wireless transactions generally are “local” and have used Cellular Market Areas (“CMAs”) as the local geographic market. In addition, it has also evaluated a transaction’s competitive effects at the national level where a transaction exhibits certain national characteristics that provide cause for concern. See, e.g., Applications of AT&T Inc., Leap Wireless International, Inc., Cricket License Co., LLC and Leap Licenseco, Inc. for Consent To Transfer Control and Assign Licenses and Authorizations, WT Docket No. 13-193, Memorandum Opinion and Order, 29 FCC Rcd 2735, 2735 ¶ 27 (WTB, IB 2014) (“AT&T-Leap Order”). 50 Antitrust authorities in the United States generally classify markets into three types: Unconcentrated (HHI < 1500), Moderately Concentrated (1500 < HHI < 2500), and Highly Concentrated (HHI > 2500). See Horizontal Merger Guidelines, U.S. Department of Justice and the Federal Trade Commission (Aug. 19, 2010), available at http://www.justice.gov/atr/public/guidelines/hmg-2010.pdf. The Commission’s initial HHI screen identifies, for further case-by-case market analysis, those markets in which, post- transaction: (1) the HHI would be greater than 2800 and the change in HHI would be 100 or greater; or (2) the change in HHI would be 250 or greater, regardless of the level of the HHI. See, e.g., AT&T-Leap Order, 29 FCC Rcd at 2753 ¶ 41 n.140; Applications of AT&T Inc. and Atlantic Tele-Network, Inc. for Consent To Transfer Control of and Assign Licenses and Authorizations, WT Docket No. 13-54, Memorandum Opinion and Order, 28 FCC Rcd 13670, 13684-85 ¶ 30 n.77 (WTB, IB 2013) (“AT&T-ATN Order”); Applications of GCI Communication Corp., ACS Wireless License Sub, Inc., ACS of Anchorage License Sub, Inc., and Unicom, Inc. for Consent To Assign Licenses to the Alaska Wireless Network, LLC, WT Docket No. 12-187, Memorandum Opinion and Order and Declaratory Ruling, 28 FCC Rcd 10433, 10450 ¶ 42 n.135 (2013) (“Alaska Wireless Order”). 51 See generally AT&T-Leap Order, 29 FCC Rcd 2735. 52 See (web) Appendix II for detailed data on the HHI by EA, available at https://www.fcc.gov/reports- research/reports?og_group_ref_target_id=1638&field_report_series_tid=1733&shs_term_node_tid_depth=All&=Apply. 53 Apart from differences in population, EAs also vary with regard to other likely determinants of market demand and facilities-based service provider costs, such as per-capita income and the age distribution of the population. Federal Communications Commission DA 15-1487 18 Chart II.C.1 Average Population-Weighted HHI Across EAs: 2004 – 2014 Chart II.C.2 2014 HHIs Plotted Against 2010 EA Population Density Federal Communications Commission DA 15-1487 19 D. Financial Indicators 1. Revenue and Average Revenue Per Unit 26. Average price metrics have been necessary and useful tools to compare broad trends in pricing in this industry, even though average metrics have always had their limitations given the longstanding variation in terms of plan characteristics and pricing for mobile voice and data.54 Total wireless service providers’ revenues, as reported by CTIA, include monthly service fees, usage-related charges, activation charges, vertical services (voice mail, enhanced calling features, and other services), out-collect roaming revenues, and data service revenues.55 In 2014, total wireless service revenue was $187.8 billion, and in contrast to previous years, was a year-over-year drop of less than one percent.56 This decline in service revenues likely indicates the beginning separation of the provider’s revenue streams into two separate components – service revenues and equipment revenues.57 Total reported prepaid revenues for 2014 equaled $24.7 billion, up 10.5 percent from $22.4 billion reported for 2013.58 27. CTIA reported an industry average measure of “Average Revenue per Reported (subscriber) Unit,” or ARPU, which is based “upon total revenues divided by the average total reported active units per survey period, divided by the number of months in the survey period,” i.e., an annualized monthly ARPU.59 As shown in 54 Different service providers have offered a variety of pricing plans for their voice and data services, with service often offered under multi-part pricing schemes and with differing non-price terms and features, such as early termination fees and the consequences of reaching usage limits. As discussed in previous Reports, it is therefore difficult to identify sources of information that track mobile wireless service prices in a comprehensive and consistent manner. Also, data on subscribership is not available at the plan level and any average price comparison implicitly assumes uniform subscribership of all plans. 55 See CTIA Year-End 2014 Wireless Indices Report, at p. 80. 56 Revenue increased by 3.0%, 4.8%, 6.2%, 9.0%, and 2.3% in 2009, 2010, 2011, 2012, and 2013, respectively. See Appendix Table II.D.i for details. 57 Analysts at MoffettNathanson Research distinguish the difference between traditional accounting for revenues and changes under Equipment Installment Plan (“EIP”) accounting, stating that “Equipment installment plans replace the traditional handset subsidy with a more transparent separation between the equipment and service components of a wireless offering. The underlying economics are not terribly different, and the monthly cost to the consumer is about the same either way.” MoffettNathan Research: “U.S. Wireless: Removing the Rose-Colored Glasses,” June 18th 2014. Under EIP, the revenue associated with the sale of a handset is fully recognized at the time of sale – regardless of whether the handset is paid for upfront or financed – accelerating revenue and therefore EBITDA and earnings. At the same time, future service revenues are reduced, lowering service ARPU and lowering EBITDA back to pre-EIP adoption levels. By contrast, in traditional subsidy plans, the full cost of the handset is recognized at the time of sale (just as in EIP accounting), but the bulk of the handset’s revenues (less the upfront price paid for the device) are recognized over a period of years, in the form of higher service revenues. The result is a near-term reduction in revenues, EBITDA (“Earnings before Interest, Taxes, Debt, and Amortization”), and margins, offset by higher service revenue in the future. 58 As detailed in Table II.C.1 above, reported company-specific revenues indicate that Verizon Wireless and AT&T had the highest service revenues in 2014, followed by Sprint, and then by T-Mobile. Beginning with its 2012 report, CTIA discontinued separately tracking and reporting wireless data service revenues. See CTIA Year-End 2013 Wireless Indices Report, at pp. 85-90 for a detailed discussion. As a result of these changes, estimates of the unit price of wireless voice and data revenues are increasingly unreliable and difficult to come by, and the Commission is no longer able to report from the CTIA data an average revenue per text message, an average revenue per megabyte, or an average voice revenue per minute. See Seventeenth Report, 29 FCC Rcd at 15328 ¶ 35. 59 See CTIA Year-End 2014 Wireless Indices Report, at p. 80. This ARPU is not equal to the “average bill” for a household or consumer as it is not equal to the bill for an “account,” which may cover several different devices, such as multiple phones (under a family plan) or multiple devices (including phones, tablets, wireless broadband modems, or other adjunct devices covered by a customer’s service plan). It assigns overall service revenue across all revenue generating devices. The total service revenues used include roaming revenues, usage fees, access, and other connection fees. See CTIA Year-End 2014 Wireless Indices Report, at pp. 1, 80. Federal Communications Commission DA 15-1487 20 Chart II.D.1 below, according to CTIA, from December 2013 to December 2014, the industry ARPU was $46.64. Chart II.D.1 also shows total service revenue, subscribers/connections and ARPU for the past 20 years, and indicates (in nominal dollars), that the ARPU is fairly stable, while revenues and connections have increased over time.60 Source: Based on CTIA Wireless Industry Indices Year-End 2014, Table 27. 2. Average Revenue Per Unit by Service Provider 28. We now present UBS estimates of ARPU and the unit price of mobile wireless broadband services. As seen in Table II.D.1, there is some variation in ARPU amongst the various national and regional wireless service providers, despite the relative stability at the industry level. Table II.D.1 below shows that between the fourth quarter of 2012 and the second quarter of 2015, all four nationwide service providers experienced a decline in ARPU. AT&T’s ARPU declined by approximately 13 percent, Verizon Wireless’s ARPU declined by approximately 9 percent, Sprint’s ARPU declined by approximately 12 percent, and T- Mobile’s ARPU declined by approximately 14 percent, respectively. The overall declines in ARPU are likely attributable to the more widespread use of EIPs, as discussed in section V. below, which enable subscribers to pay for equipment via installment payments.61 Regional service providers such as US Cellular experienced a slight increase in ARPU during this time. 60 As discussed in the Seventeenth Report, while the ARPU metric remains the best such measure currently used by industry and financial analysts, its consistent estimation has become more difficult. Industry and financial analysts have had to make additional assumptions and begun to estimate a new, normalized version of ARPU, dividing overall reported service revenues by the average number of connections for the period. See Seventeenth Report, 29 FCC Rcd at 15328-29 ¶ 36. 61 See UBS Investment Research. UBS US Wireless 411 Version 55, pp.7-9. 0 50 100 150 200 250 300 350 400 Year Chart II.D.1 Total Service Revenues, Subscribers/Connections, ARPU, 1993 - 2014 Total Annual Service Revenue ($ billions) Reported Subscribers (millions) ARPU ($) Federal Communications Commission DA 15-1487 21 Table II.D.1 ARPU Estimates of Publicly Traded Facilities-Based Mobile Wireless Service Providers 4th Quarter 2012 – 2nd Quarter 2015 Nationwide Providers 4Q12 4Q13 4Q14 2Q15 AT&T $ 46.94 $ 47.58 $ 42.04 $41.07 Verizon Wireless $ 47.57 $ 47.50 $ 45.52 $43.38 Sprint $ 43.37 $ 44.83 $ 40.44 $38.03 T-Mobile $ 40.24 $ 36.91 $ 35.56 $34.77 Regional/Rural Providers 4Q12 4Q13 4Q14 2Q15 US Cellular $ 50.89 $ 50.21 $ 53.58 $51.27 MetroPCS $ 40.86 * * * Leap Wireless $ 40.69 $ 45.55 * * NTELOS $ 52.78 $ 54.11 $ 52.35 $48.91 Cincinnati Bell $ 43.28 $ 41.35 $ 39.87 * Source: UBS Investment Research. UBS US Wireless 411 Version 51, Figure 36. UBS US Wireless 411 Version 57, Figure 53. 3. Wireless Telephone Services Consumer Price Index 29. The Consumer Price Index (“CPI”) is a measure of the average change over time in the prices paid by consumers for a fixed market basket of consumer goods and services.62 As documented in previous Reports, two different pricing indicators – the Wireless Telephone Services CPI and the per-minute price of voice service – show that mobile wireless prices have declined significantly since the launch of Personal Communications Service (“PCS”) service in the mid-1990s.63 However, given the shift in mobile voice service plans away from a defined number of monthly minutes, as discussed above, there is no simple way to calculate a per-minute price for such services, and thus, this discussion focuses on the CPI. 30. The wireless telephone services’ component of the CPI (Wireless Telephone Services CPI) is published by the U.S. Department of Labor’s Bureau of Labor Statistics (“BLS”) on a national basis.64 According to CPI data, the price (in constant dollars) of wireless service has continued to decline. From December 2013 to December 2014, the annual Wireless Telephone Services CPI decreased by 2.1 percent while the overall CPI increased by 1.6 percent and the Telephone Services CPI fell by 0.4 percent.65 The Wireless Telephone Services CPI has steadily declined since 2010 following an unchanged Wireless Telephone Services CPI in 2009 and a series of much smaller declines in the period from 2002 to 2008. Since December 1997, the Wireless Telephone Services CPI has declined nearly 45 percent, while the overall CPI has increased by approximately 36 percent. 62 The basket of goods includes over 200 categories, such as food and beverages, housing, apparel, transportation, medical care, recreation, education, and communications. The CPI allows consumers to compare the price of the basket of goods and services this month with the price of the same basket a month or a year ago. 63 See, e.g., Sixteenth Report, 28 FCC Rcd at 3876-77 ¶ 267. 64 Starting in December 1997, the basket included a category for cellular/wireless telephone services. All CPI figures discussed above were taken from BLS databases. See BLS website, available at http://www.bls.gov. The index used in this analysis, the CPI for All Urban Consumers (“CPI-U”), represents about 87% of the total U.S. population. See Bureau of Labor Statistics, “Consumer Price Index: Frequently Asked Questions,” available at http://www.bls.gov/cpi/cpifaq.htm. The Cellular CPI includes charges from all telephone companies that supply “cellular telephone services,” which are defined as “domestic personal consumer phone services where the telephone instrument is portable and it sends/receives signals for calls by wireless transmission.” This measure does not include business calls, telephone equipment rentals, portable radios, and pagers. While the CPI-U is urban-oriented, it does include expenditure patterns of some of the rural population. Information submitted by companies for the CPI is provided on a voluntary basis. 65 See Appendix Table II.D.ii for details. Federal Communications Commission DA 15-1487 22 4. Profitability Metrics 31. One measure of competition in the mobile wireless marketplace is the relative profitability of competitors. In the absence of the data necessary to estimate economic profits, accounting profits can instead be estimated using various metrics available to wireless industry observers. One such accounting profits metric, based on company data reported to the Securities and Exchange Commission (“SEC”), is EBITDA (“Earnings before Interest, Taxes, Debt, and Amortization”), which equals accounting profits before deducting interest expenses, corporate income taxes, depreciation, and amortization. In 2014, as shown in Table II.D.2 below, out of the nationwide facilities-based service providers, EBITDA per subscriber ranged from a low of $9.20 (T- Mobile) to a high of $22.67 (Verizon Wireless). Table II.D.2 Annual EBITDA per Subscriber ($/month), 2012 – 1st Half 2015 Top 5 Mobile Wireless Service Providers 2012 2013 2014 1st Half 2015 Verizon Wireless 22.21 23.56 22.67 24.52 AT&T 18.64 19.55 18.39 19.18 Sprint 6.11 7.53 9.14 11.03 T-Mobile 12.09 10.08 9.20 9.37 US Cellular 11.51 7.34 6.01 11.53 Source: UBS Investment Research. UBS US Wireless 411 Version 51, Figure 47. UBS US Wireless 411 Version 55, Figure 59. Annual figures are calculated by taking the average of each quarter for each year. 32. A second indicator of mobile wireless segment profitability is the EBITDA margin, which expresses EBITDA as a percentage of service revenue.66 Standardizing EBITDA by service revenues facilitates cross-provider comparisons. The EBITDA margin of a number of the publicly reported mobile service providers for the past three years is shown in Chart II.D.2. At year-end 2014, the EBITDA margin of the top four nationwide service providers ranged from approximately 15 percent for Sprint to approximately 42 percent for Verizon Wireless. As of the second quarter of 2015, the EBITDA margin of the top four nationwide service providers ranged from around 30 percent for Sprint and T-Mobile to around 49 percent for AT&T, and 56 percent for Verizon Wireless. 66 The EBITDA margin is equal to EBITDA divided by total revenue. Because EBITDA excludes depreciation and amortization, the EBITDA margin may provide a cleaner view of a company's core profitability. Federal Communications Commission DA 15-1487 23 Chart II.D.2 Reported EBITDA Margins (%) for Selected Publicly Traded Facilities-Based Wireless Service Providers, 2012 – 2nd quarter 2015 Source: UBS Investment Research. UBS US Wireless 411 Version 51, Fig. 46. UBS US Wireless 411 Version 57, Fig. 58. III. OVERALL MOBILE WIRELESS INDUSTRY METRICS 33. In this section, we discuss the current market trends in the mobile wireless marketplace and provide additional analysis highlighting specific changes that have occurred over the last year. Specifically, this section examines such indices as the number of connections and distribution of subscribers by geography and by demographics. Further, it analyzes the extent of mobile wireless and LTE broadband coverage, including by number of available service providers, and provides a comparison of rural versus non-rural markets. A. Network Coverage 34. The initial analysis in this section is based on a centroid methodology applied to U.S. census blocks67 overlaid on service provider coverage maps provided to the Commission through a contract with Mosaik Solutions.68 If the center point, or centroid, of a census block is within the coverage boundary of a Mosaik map, 67 A census block is the smallest geographic unit for which the Census Bureau tabulates decennial census data. There are 11,166,336 blocks designated in the 2010 Census, and they range in population from zero to several hundred. See U.S. Census Bureau, “2010 Census Summary File 1 – 2010 Census of Population and Housing, Technical Documentation” at p. 21 (March 2010), available at http://www.census.gov/prod/cen2010/doc/sf1.pdf#page=504. 68 Mosaik Solutions is an independent consulting firm that tracks coverage footprints of mobile voice and mobile data networks and provides data to the Commission under contract on facilities-based service providers in the form of coverage boundary maps based on the coverage boundaries provided to them by mobile wireless network operators. See Mosaik, “About Us,” available at http://www.mosaik.com/about-us/. 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 Verizon Wireless 46.3 49 50 41.4 50.4 49.8 51.1 47 52.14 50.29 49.5 42 55.8 56.1 AT&T 42.3 45.8 41.6 29.1 43.2 42.4 42 37.4 45.92 42.64 43.1 36.7 45.3 48.5 Sprint 14.6 17.8 15.3 8.8 19.2 17.6 16.2 14.4 25.34 25.30 19.9 15.4 25.6 31.7 T-Mobile 28.7 30.5 28.8 24 28.9 24.7 26.2 24 20.39 26.46 23.7 29.8 23.9 29.6 Metro PCS 27.2 32.1 28.9 31.9 22.6 41.1 41.5 27.9 Leap (Cricket) 16.1 21.8 21.4 18 17 24.8 18 20.8 17.4 21.5 11.1 9.1 US Cellular 22.4 22.8 19.7 13.5 20.4 19.9 18.3 -7.8 9.27 11.19 11.2 8.2 20.2 19.7 -10 0 10 20 30 40 50 60 P er ce n ta g e Federal Communications Commission DA 15-1487 24 then we consider that block to be “covered” by that service provider and/or technology.69 As we noted in earlier Reports, this methodology has the potential to overstate coverage in certain blocks.70 We then aggregate the population and land area of the covered census blocks to generate our coverage estimates. We note that these coverage estimates represent deployment of mobile networks and do not indicate the extent to which providers affirmatively offer service to residents in the covered areas. While recognizing therefore that this analysis likely overstates the coverage experienced by consumers because of limitations in Mosaik data, we find that this analysis is useful because it provides a general baseline that can be compared over time across network technologies, and service providers.71 35. We use the same centroid methodology when presenting below some summary information based on Form 477 data, which contains certain specific information on deployment at a detailed geographic level.72 We note, therefore, that coverage estimates based on Form 477 data are subject to similar methodological limitations as just described, and consequently have the potential to overstate coverage. We note that moving forward, and as also indicated in section VI. below, we anticipate that Form 477 data will be our primary source for the analysis of overall mobile wireless coverage as well as coverage by individual service providers in the mobile wireless marketplace. We first present our overall mobile wireless coverage estimates of the percentage of the U.S. population, land area, and road miles covered by a certain number of facilities-based service providers.73 We then present estimated LTE mobile broadband coverage, which is now the baseline industry standard for the 69 The centroid method overlays the geographic polygons showing wireless coverage onto a map of census blocks. The centroid method codes a census block as covered if the calculated center point (the “centroid”) of the census block is within the coverage polygon. If a centroid is covered, then all of the population and land area in the corresponding census block is also coded as covered. We note that in some cases the calculated center point may lay outside of the boundaries of a census block. In these cases, the centroid will be identified as the point inside the census block nearest to the calculated center point. See Seventeenth Report, 29 FCC Rcd at 15332-33 ¶ 45. 70 See, e.g., Seventeenth Report, 29 FCC Rcd at 15332-33 ¶ 45. These estimates likely overstate the coverage actually experienced by consumers because Mosaik reports advertised coverage as reported to it by service providers, each of which uses a different definition or determination of coverage. The data do not expressly account for factors such as signal strength, bit rate, or in-building coverage, and may convey a false sense of consistency across geographic areas and service providers. Further, we note that an analysis of coverage at the nationwide level will mask regional disparities in coverage. See id. 71 See Seventeenth Report, 29 FCC Rcd at 15332-33 ¶ 45. 72 Currently, Form 477 collects data from facilities-based service providers of (1) internet service with information transfer rates exceeding 200 kbps in at least one direction; and (2) mobile service to at least one subscriber. This excludes service providers of terrestrial wireless “hot spot” services, like local-area Wi-Fi or Wi-Fi within public places, but includes service providers whose only customers are served via reseller of mobile services. Facilities-based service providers of mobile wireless service submitted polygons in an ESRI shapefile format representing geographic coverage nationwide (including U.S. territories) for each transmission technology (e.g., EV-DO, WCDMA, HSPA+, LTE, WiMAX) deployed in each frequency band (e.g., 700 MHz, Cellular, AWS, PCS, BRS/EBS). See FCC website, “FCC Encyclopedia, Form 477 Orders and Releases,” available at https://www.fcc.gov/encyclopedia/form-477-orders-and-releases. In addition, service providers submit information on the geographic areas in which users should expect to receive the minimum speed advertised by the provider for the used spectrum and deployed technologies. The deployment data, with the exception of certain spectrum and speed information associated with the coverage areas, will be made public while the provider-specific subscription data will remain confidential. See Modernizing the FCC Form 477 Data Program, WC Docket No. 11-10, Report and Order, 28 FCC Rcd 9887, 9921-23 ¶¶ 79-83 (2013). See also FCC website, “Changes to the Form 477 Data Collection in 2014,” available at https://www.fcc.gov/encyclopedia/changes-coming-form-477-data-collection. 73 We note that service providers often offer coverage outside of their network coverage areas through roaming arrangements which allow their customers to automatically receive service from other providers’ networks when they are in areas that are covered by their roaming partners’ networks but not by their own network. In contrast to the purchase of capacity wholesale to provide resale or MVNO services, a provider uses roaming services to market extended coverage to consumers residing within the provider’s network coverage area, but not to acquire customers where a service provider does not have network coverage. See Appendix Table III.A.i and Table III.A.ii for more details of estimated overall mobile wireless coverage. Federal Communications Commission DA 15-1487 25 marketing of mobile broadband service, using the same categories.74 We note that LTE deployment does not necessarily result in specific guaranteed speeds for the consumer. 1. Overall Mobile Wireless Network Coverage 36. Subject to the limitations just described, the available data suggest quite extensive mobile wireless coverage. In looking at comparable past data, we note that the percentage of the population living in a census block with mobile wireless coverage by at least one or two providers has not changed significantly since January 2012.75 As of July 2015, while more than 90 percent of the U.S. population lived in census blocks with coverage by at least four providers, these census blocks accounted for only approximately 31 percent of the total land area of the United States, and approximately 55 percent of U.S. road miles. Chart III.A.2 presents the results from the same analysis using Form 477 data. According to both sources, more than 90 percent of the population is covered by at least four service providers, but coverage of road-miles and of land area is much more limited. Chart III.A.1 Estimated Mobile Wireless Coverage by Census Block: Mosaik, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 74 For coverage estimates of mobile service provided using any of the following 3G or 4G technologies: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX, see Appendix III, Chart III.A.i, and Tables III.A.iii and III.A.iv. As of July 2015, mobile wireless coverage of 3G and better technologies is not materially different than coverage that also includes 2G technologies. 75 See Seventeenth Report, 29 FCC Rcd at 15333-34 ¶ 47. The percentages of population located in census blocks with coverage by a certain number of mobile wireless or mobile LTE broadband providers does not necessarily mean that those service providers offered service to residents in the census block. In addition, we emphasize that a service provider reporting mobile wireless or mobile LTE broadband coverage in a particular census block may not provide coverage everywhere in the census block. For both these reasons, the number of providers in a census block does not necessarily reflect the number of choices available to a particular individual or household, and does not purport to measure competition. 9.5% 30.6% 49.6% 68.7% 78.0% 16.9% 55.3% 75.2% 91.7% 97.1% 19.4% 91.7% 97.2% 99.6% 99.9% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile Wireless Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Federal Communications Commission DA 15-1487 26 Chart III.A.2 Estimated Mobile Wireless Coverage by Census Block: Form 477, Dec. 201476 Note: The coverage calculation methodology has certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 37. In this Report, we have included a service provider if it has a market share above a particular threshold, and we have made estimates based on two alternative thresholds. Specifically, to estimate the number of service providers serving a CMA, we include a service provider if it has a greater than two percent market share (or alternatively, a five percent market share which provides greater assurance of a meaningful choice for consumers) of mobile wireless connections based on NRUF data. Table III.A.1 presents the data for December 2011 and December 2014, and shows that since 2011, based on the five percent market share threshold, the number of CMAs with three service providers has decreased, while the number of CMAs with four service providers has substantially increased.77 There has been a decrease in the number of CMAs with at least five service providers based on the five percent market share threshold primarily due to increased industry consolidation, as noted above. 76 In order to calculate coverage based on Form 477 data, the same centroid methodology is used as in our calculations of coverage based on Mosaik data. As previously noted, the use of the centroid methodology has certain limitations that likely result in an overstatement of the extent of mobile wireless or mobile LTE broadband coverage. Further, “although most census blocks are small, some can be large, particularly in low-density rural areas, and … coverage at the centroid might result, incorrectly, in the entirety of those large areas being deemed served.” Connect America Fund, et al., WC Docket No. 10-90, et al., Report and Order and Further Notice of Proposed Rulemaking, 26 FCC Rcd 17663, 17787 ¶ 344 (2011). 77 Because NRUF includes data on the number of telephone numbers that have been assigned to end-user devices by mobile wireless providers, this analysis does not include providers whose data-only devices are not assigned a mobile telephone number. 9.2% 30.1% 49.5% 71.0% 80.5% 15.8% 53.4% 73.1% 91.1% 97.0% 19.2% 91.2% 96.6% 99.6% 99.9% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile Wireless Coverage (%)N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (D ec em b er 2 0 1 4 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Source: Form 477 Dec. 2014, 2010 Census Federal Communications Commission DA 15-1487 27 Table III.A.1 Estimated Mobile Wireless Service Providers Offering Service by CMA, Excluding Territories Two Percent Market Share Threshold Five Percent Market Share Threshold Number of Providers Offering Service Anywhere in a CMA Number of CMAs Total CMAs (percent) Number of CMAs Total CMAs (percent) 2011 2014 2011 2014 2011 2014 2011 2014 Total for U.S., excluding territories 716 716 100.0% 100.0% 716 716 100.0% 100.0% 1 provider 1 0 0.1% 0.0% 2 2 0.3% 0.3% 2 providers 51 75 7.1% 10.5% 120 159 16.8% 22.2% 3 providers 152 121 21.2% 16.9% 213 167 29.7% 23.3% 4 providers 208 377 29.1% 52.7% 246 331 34.4% 46.2% 5 or more providers 304 143 42.5% 20.0% 135 57 18.9% 8.0% Source: Based on December 2011 and December 2014 NRUF data. 2. LTE Mobile Broadband Coverage 38. Chart III.A.3 presents LTE mobile broadband coverage as of July 2015, based on Mosaik data.78 As noted above, LTE deployment does not necessarily result in specific guaranteed speeds for the consumer. In July 2015, approximately 92 percent of the population was covered by three or more service providers, while approximately 82 percent of the population was covered by four or more service providers. Chart III.A.3 also shows that LTE deployment tends to focus on the larger population centers. While over 80 percent of the U.S. population lived in census blocks with LTE coverage by at least four service providers in July 2015, these census blocks only accounted for approximately 34 percent of road miles, and approximately 15 percent of the total land area of the United States. The results are similar based on the Form 477 data: According to both Mosaik and Form 477 data, while over 80 percent of the population is covered by at least four service providers, coverage of road-miles and of land area is much more limited. 78 For coverage estimates of mobile service including any of the following 3G or 4G technologies: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX, see Appendix III, Chart III.A.i and Tables III.A.iii and III.A.iv, and III.A.v. Federal Communications Commission DA 15-1487 28 Chart III.A.3 Estimated LTE Coverage in the U.S. by Census Block: Mosaik, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 3. Rural/Non-Rural Comparisons 39. While the Communications Act does not include a statutory definition of what constitutes a rural area, since its 2004 Report and Order concerning the deployment of wireless services in less populated areas, the Commission has used a “baseline” definition of rural as a county with a population density of less than 100 people per square mile.79 By this definition, approximately 56 million people, or approximately 18 percent of the U.S. population, live in rural counties, based on 2010 US Census data. These counties comprise approximately 3 79 See Facilitating the Provision of Spectrum-Based Services to Rural Areas and Promoting Opportunities for Rural Telephone Companies To Provide Spectrum-Based Services, WT Docket No. 02-381, WT Docket No. 01-14, WT Docket No. 03-202, Report and Order and Further Notice of Proposed Rule Making, 19 FCC Rcd 19078, 19086-88 ¶¶ 10-12 (2004) (“2004 Report and Order”) (“We recognize, however, that the application of a single, comprehensive definition for ‘rural area’ may not be appropriate for all purposes. . . Rather than establish the 100 persons per square mile or less designation as a uniform definition to be applied in all cases, we instead believe that it is more appropriate to treat this definition as a presumption that will apply for current or future Commission wireless radio service rules, policies and analyses for which the term ‘rural area’ has not been expressly defined. By doing so, we maintain continuity with respect to existing definitions of ‘rural’ that have been tailored to apply to specific policies, while also providing a practical guideline”). Id., 19 FCC Rcd at 19087-88 ¶ 12. 2.1% 14.5% 27.1% 48.8% 67.7% 4.5% 34.3% 51.7% 75.6% 90.7% 6.7% 82.2% 91.5% 97.8% 99.6% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile LTE Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e J u ly 2 0 1 5 % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Federal Communications Commission DA 15-1487 29 million square miles, or approximately 84 percent of the geographic area of the United States.80 We use this same definition of rural to analyze coverage by census block in rural versus non-rural areas in this Report. (i) Mobile Wireless Network Coverage 40. Chart III.A.4 shows that the difference between rural and non-rural voice coverage by census block was generally more pronounced as the number of service providers increased.81 While approximately 96 percent of the population living in non-rural areas was covered by four service providers, only approximately 63 percent of the rural population was covered. The recent acquisitions of urban-focused service providers, MetroPCS and Leap Wireless (“Cricket”) by T-Mobile and AT&T, respectively, are reflected in the numbers for areas covered by five or more network providers. In addition, Appendix Chart III.A.ii presents mobile wireless coverage of rural and non-rural road miles as of July 2015 and shows that non-rural road miles are consistently covered by more service providers than rural road miles.82 Chart III.A.4 Estimated Mobile Wireless Voice Population Coverage by Census Block in Rural vs. Non-Rural Areas: Mosaik, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 80 Based on 2010 Census data (includes the population of Puerto Rico). 81 See also Appendix Tables III.A.vi and III.A.vii. In addition, the results from Form 477 are similar – the difference between rural and non-rural coverage is generally more pronounced, the higher the number of service providers. 82 See also Appendix Tables III.A.vi and III.A.vii. 24.3% 62.6% 84.0% 97.0% 99.4% 16.4% 95.5% 99.1% 99.9% 100.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile Wireless Voice Coverage N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Non-Rural POPs % of U.S. Rural POPsSource: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 30 (ii) Mobile LTE Network Coverage 41. Chart III.A.5 shows that the coverage gap as of July 2015 between rural and non-rural population living in census blocks increased as the number of service providers increased.83 Considering LTE coverage by at least three service providers, Chart III.A.5 shows that while almost 98 percent of the non-rural population was covered by at least three service providers, approximately 65 percent of the rural population had the same network coverage. Considering coverage by at least four service providers, Chart III.A.5 shows that while approximately 92 percent of non-rural America was covered, only approximately 41 percent of rural America was covered. Our analysis of Form 477 yields similar results. While approximately 91 percent of non-rural America was covered by at least four service providers, only approximately 35 percent of rural America was covered.84 In addition, Appendix Chart III.A.iii presents mobile wireless LTE coverage by census block of rural and non-rural road miles in July 2015, and shows that non-rural road miles are consistently covered by more service providers than rural road miles.85 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 83 See also Appendix Tables III.A.viii and III.A.X.xix. 84 Staff estimates based on Form 477 data, Dec. 2014. 85 See also Appendix Tables III.A.viii and III.A.xix. 8.3% 40.7% 64.8% 90.0% 97.9% 6.3% 91.9% 97.8% 99.7% 99.9% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% 5 or More 4 or More 3 or More 2 or More 1 or More Mobile LTE Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e D ec 2 0 1 4 Chart III.A.5 Estimated Mobile LTE Coverage in Non-Rural and Rural Areas by Census Block; Mosaik, July 2015 % of Non-Rural U.S. Population % of Rural U.S. Population Federal Communications Commission DA 15-1487 31 4. Mobile Wireless and LTE Broadband Coverage by Income Levels 42. Charts III.A.7a-c below show how the number of facilities-based mobile wireless service providers that have mobile wireless and LTE coverage in a census tract varies by median household income levels.86 As of July 2015, the charts show that the average number of mobile wireless providers with coverage in census tracts was around four, no matter the ruralness of the census tract, nor the median income level. The charts reveal a very different picture for the number of service providers with LTE coverage by census tract. For households with a median income of less than $25,000, the average number of service providers with LTE coverage in non-rural areas was 4.14, as compared to 3.42 in rural areas, whereas for households with a median income of $25,000 to $50,000, the average numbers were 4.00 and 3.71, respectively. In addition, the average number of service providers with LTE coverage further declined to 3.22 in extremely rural areas for households with a median income level of less than $25,000. We further note that for households with median income of $75,000 to $100,000, there is a difference between LTE coverage in rural versus extremely rural areas: while the average number of providers with coverage is approximately four in rural areas, it drops to approximately three in extremely rural areas. 86 The percentages of population located in census tracts with a certain number of mobile wireless or mobile LTE broadband providers represent network coverage, which does not necessarily mean that they offered service to residents in the census block. In addition, we emphasize that a service provider reporting mobile wireless or LTE broadband coverage in a particular census tract may not provide coverage everywhere in the census tract. For both these reasons, the number of service providers in a census tract, or by income level does not necessarily reflect the number of choices available to a particular individual or household at a certain income level, and does not purport to measure competition. In addition, calculations based on Mosaik data on coverage, while useful for measuring developments in mobile wireless coverage, have certain limitations that likely result in an overstatement of the extent of mobile wireless and mobile LTE broadband coverage. Data on median household income are based on the U.S. Census Bureau’s American Community Survey (“ACS”) 2009- 2013. The analysis is done on a census tract, rather than census block, basis because the smallest geographic area for which median household income data is available is census tracts. These data do not allow for an analysis of adoption rates for mobile wireless or mobile LTE broadband services. 4.37 4.29 4.25 4.17 4.12 4.09 4.054.14 4.00 4.04 4.03 4.02 4.01 4.01 0.00 1.00 2.00 3.00 4.00 5.00 $0-$25K $25K-$50K $50K-$75K $75K-$100K $100K-$125K $125K-$150K $150K+N u m b er o f W ir el es s P ro v id er s an d L T E P ro v id er s Median Household Income by Census Tract Chart III.A.7a Average Number of Mobile Wireless and LTE Providers in Census Tracts by Median Household Income in Non-Rural Areas with Population Density more than 100 People per Square Mile, July 2015 Ave Number of Mobile Wireless Providers July 2015 Ave Number of LTE Providers 2015 Federal Communications Commission DA 15-1487 32 Source: Data on median household income by census tract are based on the U.S. Census Bureau’s American Community Survey (“ACS”) 2009-2013. Data on the number of service providers are from Mosaik, July 2015. It is important to note that the number of mobile wireless or mobile LTE broadband service providers in a census tract represents network coverage, which does not necessarily mean that they offered service to any or all the residents in the census tract. In addition, we emphasize that a service provider reporting mobile wireless or LTE broadband coverage in a particular census tract may not provide coverage everywhere in the tract. 4.17 4.48 4.62 4.51 4.47 4.25 4.00 3.42 3.71 4.00 4.01 3.97 4.08 4.00 0.00 1.00 2.00 3.00 4.00 5.00 $0-$25K $25K-$50K $50K-$75K $75K-$100K $100K-$125K $125K-$150K $150K+ N u m b er o f W ir el es s P ro v id er s an d L T E P ro v id er s Median Household Income by Census Tract Chart III.A.7b Average Number of Mobile Wireless and LTE Providers in Census Tracts by Median Household Income in Rural Areas with Population Density between 20 and 100 People per Square Mile, July 2015 Ave Number of Mobile Wireless Providers July 2015 Ave Number of LTE Providers 2015 4.45 4.43 4.61 4.13 4.69 4.33 4.00 3.22 3.30 3.42 3.09 4.00 3.67 4.00 0.00 1.00 2.00 3.00 4.00 5.00 $0-$25K $25K-$50K $50K-$75K $75K-$100K $100K-$125K $125K-$150K $150K+ N u m b er o f W ir el es s P ro v id er s an d L T E P ro v id er s Median Household Income by Census Tract Chart III.A.7c Average Number of Mobile Wireless and LTE Providers in Census Tracts by Median Household Income in Extremely Rural Areas with Population Density Less Than 20 People per Square Mile, July 2015 Ave Number of Mobile Wireless Providers July 2015 Ave Number of LTE Providers 2015 Federal Communications Commission DA 15-1487 33 B. Connections and Subscribers 1. Connections and Subscribers by Geography 43. To better understand the number of connections across geographic areas, for this Report, we have estimated penetration rates (the number of mobile wireless connections per 100 people), using NRUF subscriber data, for the 172 EAs of the United States. As discussed above, we use EAs as the geographic unit for measuring the level of concentration in the mobile wireless services marketplace in order to maintain continuity with past Reports and to ensure that we do not compromise the confidential information contained in the NRUF data.87 Regional penetration rates for the 172 EAs range from 87 percent in Wausau, Wisconsin to 151 percent in Wichita, Kansas.88 The nationwide penetration rate based on NRUF data now exceeds 100 percent, meaning that the number of connected devices exceeds the total population of the United States, and the penetration rate was at least 100 percent in 121 of the 172 EAs at the end of 2014.89 2. Connections and Subscribers by Demographics 44. Several socio-economic and demographic factors such as household income and age are correlated with overall mobile wireless subscription rates as well as smartphone subscription rates. Based on September 2015 survey data from ComScore Mobilens,90 Chart III.B.1 below shows that mobile wireless subscribers overall, and smartphone subscribers in particular, are in higher income brackets. For example, approximately 25 percent of the population live in households with an annual income of less than $25,000, but only approximately 18 percent of mobile wireless users and approximately 15 percent of smartphone users are in this bracket. Conversely, approximately 22 percent of the population live in households with an annual income over $100,000, but approximately 28 percent of mobile wireless subscribers and approximately 31 percent of smartphone subscribers are in this income bracket. Further, we note that more postpaid users are in a higher income bracket, while the converse is true for prepaid subscribers. 45. As shown in Chart III.B.2 below, the ComScore data also allow the presentation of the composition of mobile users by age. While the general adoption of mobile wireless devices is fairly evenly distributed among various age groups, smartphone adoption is more concentrated in younger age groups. For example, as of September 2015, adults ages 18 to 44 comprise approximately 46 percent of all mobile wireless subscribers, but make up approximately 54 percent of smartphone subscribers, while adults 55 or over represent approximately 29 percent of all mobile wireless subscribers, but only approximately 21 percent of smartphone subscribers. 87 See section II.D. supra. 88 See (web) Appendix III for detailed data on penetration rates by EA, available at https://www.fcc.gov/reports- research/reports?og_group_ref_target_id=1638&field_report_series_tid=1733&shs_term_node_tid_depth=All&=Apply. 89 According to the U.S. Census Bureau, the combined population of the 50 states, the District of Columbia, and Puerto Rico, as of July 1, 2014, was estimated to be 322.4 million. See U.S. Census Bureau website, “American FactFinder,” available at http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=PEP_2014_PEPANNRES&src=pt. We note that if NRUF is used to calculate a mobile wireless penetration rate (of a population), that penetration rate is overstated due to the number of individuals who have more than one mobile wireless device. 90 Survey data based on ComScore MobiLens, September 2015. ComScore MobilLens U.S. data are derived from a monthly survey of over 13,000 respondents ages 13 and older who are recruited to represent U.S. Census demographics. The total universe size is estimated from data provided by CTIA and comScore’s monthly subscriber studies. Race data are found at the U.S. Census Bureau State and County Quick Facts 2013 website, available at http://quickfacts.census.gov/qfd/states/00000.html. Income data are found in the “Income, Poverty, and Health Insurance Coverage in the United States: 2012 Current Population Reports. United States Census Bureau” (Issued Sept. 2013), Table A-1: “Households by Total Money Income, Race, and Hispanic Origin of Householder,” available at http://www.census.gov/prod/2013pubs/p60-245.pdf. Federal Communications Commission DA 15-1487 34 Source: ComScore, MobiLens Audience Profile September 2015, and U.S. Census Bureau. Source: ComScore MobiLens, 3-Month Average, September 2015. 18.2% 15.4% 11.9% 28.1% 21.1% 20.1% 18.6% 25.5% 19.0% 19.1% 20.0% 17.8% 13.7% 14.5% 15.1% 11.3% 28.0% 30.9% 34.4% 17.3% U.S. Mobile Phone Subscribers U.S. Smartphone Subscribers Postpaid Subscribers Prepaid Subscribers T o ta l S u b sc ir b er s Chart III.B.1 Total Mobile Subscribers, Smartphone Subscribers, and Prepaid/Postpaid Subscribers by Income Bracket, September 2015 $100k+ $75k to < $100k $50k to < $75k $25k to < $50k less than $25k * CTIA: 355.4 million (12/14) * NRUF: 357.1 million (12/14) 7.2% 7.6% 12.8% 15.3% 16.9% 20.2% 16.2% 18.2% 17.7% 17.6% 15.3% 12.7% 14.0% 8.4% U.S. Mobile Phone Subscribers U.S. Smartphone Subscribers T o ta l S u b sc ri b er s Chart III.B.2 Total Mobile Subscribers Compared to Smartphone Subscribers by Age, September 2015 13-17 yrs old 18-24 yrs old 25-34 yrs old 35-44 yrs old 45-54 yrs old 55-64 yrs old 65+ yrs old Federal Communications Commission DA 15-1487 35 IV. INPUT MARKETS 46. Mobile wireless service providers employ a combination of inputs to provide mobile wireless services to their customers. These inputs include electromagnetic spectrum to transmit signals between base stations and end users’ devices, as well as non-spectrum inputs such as cellular base stations and towers to carry transmissions. Further, backhaul, which routes voice and data traffic from base stations for onward transmission and may use spectrum or wireline resources, is an additional input required for the provision of mobile service. In this section, we first discuss the critical role of spectrum as an input in the provision of mobile wireless services. Next, we summarize the Commission’s policies to facilitate the use of spectrum and then provide information on service providers’ current spectrum holdings. Lastly, we provide an analysis of non-spectrum inputs. A. Spectrum 1. Importance of Spectrum for the Provision of Mobile Wireless Services 47. As the Commission has found, spectrum is a critical input in the provision of mobile wireless services and affects if and when existing service providers and potential entrants will be able to expand capacity or deploy networks.91 Incumbent service providers may need additional spectrum to increase their coverage or capacity, while new entrants need access to spectrum to enter a geographic area.92 In addition, increasing consumer demand for mobile broadband is increasing service providers’ need for spectrum at an unprecedented rate and this is projected to grow further.93 Spectrum bands vary in their propagation characteristics, which has implications for how spectrum is deployed.94 Spectrum below 1 GHz (“low-band spectrum”) has certain propagation advantages for network deployment over long distances, while also reaching deep into buildings and urban canyons, while spectrum above 1 GHz (“high-band spectrum”) allows for the better transmission of large amounts of information.95 As service providers deploy next-generation mobile networks, the engineering properties and deployment capabilities of the mix of particular spectrum bands have become increasingly important, particularly as multi-band phones allow service providers to take advantage of these different properties.96 48. Competition in the mobile wireless marketplace will be better promoted by multiple service providers having the opportunity to access both low-band spectrum that can provide coverage and in-building penetration, as well as high-band spectrum that can provide the increased throughput for mobile broadband applications.97 Service providers holding a mix of low- and high-band spectrum licenses have greater flexibility 91 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6134 ¶ 2. See also Seventeenth Report, 29 FCC Rcd at 15355 ¶ 89. 92 See Seventeenth Report, 29 FCC Rcd at 15355 ¶ 89. 93 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6134 ¶ 2. 94 Service providers deploy their spectrum bands differently depending on the nature of the service, geography, density, or other factors in their network build-out. See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6149-54 ¶¶ 31-40; Seventeenth Report, 29 FCC Rcd at 15355-57 ¶¶ 90-92. 95 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6135 ¶ 3; Seventeenth Report, 29 FCC Rcd at 15356 ¶ 90. In this sense, low-band spectrum may be thought of as “coverage” spectrum, and high-band spectrum may be thought of as “capacity” spectrum. We note that there is significantly less low-band spectrum than high-band spectrum that is suitable and available for the provision of mobile telephony/broadband services. See Seventeenth Report, 29 FCC Rcd at 15356 ¶ 90. 96 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6144 ¶ 18; Seventeenth Report, 29 FCC Rcd at 15356-57 ¶¶ 91-92. 97 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6157 ¶ 47; Seventeenth Report, 29 FCC Rcd at 15356 ¶ 91. Federal Communications Commission DA 15-1487 36 and are better able to optimize network costs for a given quality level.98 Service providers without access to low- band spectrum would have to rely on less efficient and cost-effective methods to increase rural and in-building coverage to serve additional customers, such as adding towers, splitting cells, or acquiring roaming rights on other networks.99 2. Facilitating Access to Spectrum 49. Recognizing the importance of spectrum in the provision of mobile wireless services, Congress, through the Communications Act, requires the Commission to implement spectrum policies that promote competition, innovation, and the efficient use of spectrum to serve the public interest, convenience, and necessity.100 Further, policies which promote and preserve competition, in turn, enable consumers to make choices among multiple service providers and lead to lower prices, improved quality, and increased innovation.101 Consistent with this statutory mandate, the Commission has established policies to make spectrum available to existing mobile service providers and potential new entrants through initial licensing, primarily by competitive bidding, and through secondary market transactions.102 The Commission generally has adopted “flexible use” policies, thereby allowing licensees to decide which services to offer and what technologies to deploy on spectrum used for the provision of mobile wireless services. a. Auctions 50. Since 1994, the Commission has conducted various auctions of spectrum licenses.103 These auctions are open to any eligible entity that submits an application and upfront payment and is found to be a qualified bidder by the Commission.104 In addition, the Commission generally provides a bidding credit – or discount – to promote participation by small businesses and rural service providers, including businesses owned by members of minority groups and women (collectively “designated entities”).105 The Seventeenth Report discusses auctions for the various frequency bands which are potentially suitable for the provision of mobile telephony/broadband service.106 Further, the Commission’s auction website provides detailed information regarding completed, ongoing, and planned auctions.107 To provide service providers with the opportunities to 98 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6163-64 ¶ 59; Seventeenth Report, 29 FCC Rcd at 15356-57 ¶ 92. 99 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6164 ¶ 60. While other cost-related factors exist, ensuring that multiple service providers are able to access a sufficient amount of low-band spectrum is a threshold requirement for extending and improving service in both rural and urban areas. See id., 29 FCC Rcd at 6135 ¶ 3. 100 47 U.S.C. § 309(j)(3)(B). 101 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6144 ¶ 17. 102 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6143-44 ¶ 17, 6167-68 ¶¶ 67-69, 6190 ¶ 135, 6193 ¶ 144, 6221-22 ¶¶ 225-27, 6223-24 ¶¶ 231-32. 103 See FCC website, “Auctions Home,” available at http://wireless.fcc.gov/auctions/default.htm?job=auctions_home. 104 See FCC website, “About Auctions,” available at http://wireless.fcc.gov/auctions/default.htm?job=about_auctions. 105 See 47 U.S.C. § 309(j)(3)(B), 309(j)(4)(D); see also 47 C.F.R. § 1.2110. Updating Part 1 Competitive Bidding Rules; Expanding the Economic and Innovation Opportunities of Spectrum Through Incentive Auctions; Petition of DIRECTV Group, Inc. and EchoStar LLC for Expedited Rulemaking to Amend Section 1.2105(a)(2)(xi) and 1.2106(a) of the Commission’s Rules and/or for Interim Conditional Waiver; Implementation of the Commercial Spectrum Enhancement Act and Modernization of the Commission’s Competitive Bidding Rules and Procedures, WT Docket No. 14-170, GN Docket No. 12-268, RM 11395, WT Docket No. 05-211, Report and Order, Order on Reconsideration of the First Report and Order, Third Order on Reconsideration of the Second Report and Order, and Third Report and Order, 30 FCC Rcd 7493 (2015) (“Part 1 Report and Order”) (modified by Erratum; rel. Aug. 25, 2015). 106 See Seventeenth Report, 29 FCC Rcd at 15311-19 ¶¶ 95-96. 107 See FCC website, “Auctions Home,” available at http://wireless.fcc.gov/auctions/default.htm?job=auctions_home. Federal Communications Commission DA 15-1487 37 better meet the rising consumer demand discussed above, the Commission has made, and is continuing to make, substantially more spectrum available for the provision of mobile wireless services. For example, in February 2014, the Commission auctioned the 10 megahertz of H Block in the 1.9 GHz Band with DISH winning each of the 176 licenses.108 In January 2015, the Commission auctioned 65 megahertz of high-band spectrum in the Advanced Wireless Services-3 (“AWS-3”) auction, generating approximately $45 billion in (gross) bids.109 A total of 31 bidding entities won spectrum in the auction, and the spectrum sold for an average of $2.71 per MHz- POP for paired spectrum and $0.52 per MHz-POP for unpaired spectrum.110 The 600 MHz Incentive Auction (“incentive auction”), which is scheduled to begin in March 2016, has the potential to make available significant amounts of low-band spectrum currently used for over-the-air television broadcasting.111 51. In the Mobile Spectrum Holdings Report and Order, the Commission concluded that, in lieu of a post-auction application of the spectrum screen to the initial licensing of spectrum to winning bidders, the Commission would determine whether a band-specific mobile spectrum holding limit is necessary, and if so, would establish an ex ante application of that limit to the competitive bidding for that band.112 The Commission declined to adopt band-specific mobile spectrum holding limits for AWS-3, emphasizing the availability of a substantial amount of comparable high-band spectrum to competitors and the significant existing holdings of multiple service providers of comparable spectrum.113 Regarding the incentive auction, however, the Commission established a market-based spectrum reserve of up to 30 megahertz in each geographic license area (Partial Economic Area) (“PEA”)114 that is designed to ensure against excessive concentration in holdings of low-band spectrum while including safeguards to ensure that all bidders bear a fair share of the cost of the incentive auction.115 52. In the incentive auction, the Commission will allow broadcasters to voluntarily participate in a “reverse auction” of the 600 MHz spectrum, which will then be made available in a “forward auction” to be used 108 See Auction of H Block Licenses in the 1915-1920 MHz and 1995-2000 MHz Bands Closes; Winning Bidder Announced for Auction 96, Public Notice, 29 FCC Rcd 2044 (2014); Service Rules for Advanced Wireless Services in the 2000-2020 MHz and 2180-2200 MHz Bands, WT Docket No. 12-70, Fixed and Mobile Services in the Mobile Satellite Service Bands at 1525-1559 MHz and 1626.5-1660.5 MHz, 1610-1626.5 MHz and 2483.5-2500 MHz, and 2000-2020 MHz and 2180-2200 MHz, ET Docket No. 10-142, Service Rules for Advanced Wireless Services in the 1915-1920 MHz, 1995-2000 MHz, 2020- 2025 MHz and 2175-2180 MHz Bands, WT Docket No. 04-356, Report and Order and Order of Proposed Modification, 27 FCC Rcd 16102 (2012) (“AWS-4 Report and Order”). 109 See Auction Of Advanced Wireless Services (AWS-3) Licenses Closes, Winning Bidders Announced For Auction 97, Public Notice, 30 FCC Rcd 630 (2015). See also Amendment of the Commission’s Rules with Regard to Commercial Operations in the 1695-1710 MHz, 1755-1780 MHz, and 2155-2180 MHz Bands, GN Docket No. 13-185, Report and Order, 29 FCC Rcd 4610 (2014) (“AWS-3 Report and Order”). 110 See Phil Goldstein, “AWS-3 Auction Results: AT&T leads with $18.2B, Verizon at $10.4B, Dish at $10B and T-Mobile at $1.8B” (Jan. 30, 2015), available at http://www.fiercewireless.com/story/aws-3-auction-results-att-leads-182b-verizon- 104b-dish-10b-and-t-mobile-18b/2015-01-30. 111 See Broadcast Auction Scheduled to Begin March 29, 2016; Procedures for Competitive Bidding in Auction 1000, Including Initial Clearing Target Determination, Qualifying to Bid, and Bidding in Auctions 1001 (Reverse) and 1002 (Forward), AU Docket No. 14-252, GN Docket No. 12-268, WT Docket No. 12-269, MB Docket No. 15-146, Public Notice, 30 FCC Rcd 8975 (2015). 112 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6192 ¶ 139. 113 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6220-21 ¶¶ 222-24. 114 We note that the geographic license areas for spectrum offered vary across the various auctions the Commission has conducted. For example, in Auction 96 (H Block), 176 licenses were offered (one for each of the 176 EAs), while in Auction 97 (AWS-3), 1,614 licenses were offered in total, at both the EA and CMA level. Finally, for the upcoming incentive auction, spectrum licenses will be offered at the PEA geographic level. 115 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6193-6219 ¶¶ 146-217. Federal Communications Commission DA 15-1487 38 for the provision of mobile wireless service, with flexible use service rules.116 The Commission stated that by offering only paired blocks in a single band, and by licensing on a PEA basis, the 600 MHz Band Plan will promote participation by both larger and smaller wireless service providers, including rural service providers, and encourage new entrants, and further will promote interoperability and international harmonization.117 The Commission established procedures necessary to carry out the incentive auction in the Auction 1000 Bidding Procedures Public Notice.118 In particular, this Public Notice adopted procedures for: an initial clearing target, allowing market forces to determine the highest and best use of spectrum on a near-nationwide basis; improved transparency for reverse and forward auction bidders; two categories of generic spectrum blocks for bidding in the clock phase of the forward auction; and the market-based spectrum reserve in the forward auction.119 Among the spectrum reserve procedures adopted, the Auction 1000 Bidding Procedures Public Notice adopted the Commission’s proposed average price and spectrum benchmarks to help ensure that winning bids for the licenses in the forward auction reflect competitive prices and return a portion of the value of the spectrum to taxpayers without reducing the amount of spectrum repurposed for new, flexible-use licenses.120 In addition, to promote participation by designated entities, the Commission adopted bidding credits for small businesses and, for the first time, a bidding credit for eligible rural service providers.121 The Commission capped the overall amount of bidding credit that an entity may receive.122 b. Secondary Markets 53. Subject to the Commission’s approval, licensees may assign and exchange licenses, in whole or in part (through partitioning and/or disaggregation), on the secondary market.123 In reviewing proposed acquisitions of spectrum through secondary market transactions, the Commission uses an initial screen to help identify for case-by-case review local markets where changes in spectrum holdings resulting from the transaction may be of particular concern.124 As set out in various transactions orders, however, the Commission has not limited its consideration of potential competitive harms solely to markets identified by its initial screen, if it 116 See generally Expanding the Economic and Innovation Opportunities of Spectrum Through Incentive Auctions, GN Docket No. 12-268, Report and Order, 29 FCC Rcd 6567 (2014) (“Incentive Auctions Report and Order”). 117 See Incentive Auctions Report and Order, 29 FCC Rcd at 6585-86 ¶ 44. 118 See generally Procedures for Competitive Bidding in Auction 1000, Including Initial Clearing Target Determination, Qualifying to Bid, and Bidding in Auctions 1001 (Reverse) and 1002 (Forward), AU Docket No. 14-252, GN Docket No. 12- 268, WT Docket No. 12-269, MB Docket No. 15-146, Public Notice, 30 FCC Rcd 8975 (2015) (“Auction 1000 Bidding Procedures Public Notice”). 119 See Auction 1000 Bidding Procedures Public Notice, 30 FCC Rcd at 8979 ¶ 2. 120 See id. 121 See Part 1 Report and Order, 30 FCC Rcd 7953. 122 See id., 30 FCC Rcd at 7544-48 ¶¶ 122-30. 123 As part of its secondary market policies, the Commission also permits mobile wireless licensees to lease all or a portion of their spectrum usage rights for any length of time within the license term and over any geographic area encompassed by the license. For a more comprehensive overview of the Commission’s secondary market policies. See Seventeenth Report, 29 FCC Rcd at 15358-60 ¶¶ 97-100. 124 See, e.g., Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6221-22 ¶ 225; Applications of AT&T Inc., E.N.M.R Telephone Cooperative, Plateau Telecommunications, Inc., New Mexico RSA 4 East Limited Partnership, and Texas RSA 3 Limited Partnership for Consent To Assign Licenses and Authorizations, WT Docket No. 14-144, Memorandum Opinion and Order, 30 FCC Rcd 5107, 5113 ¶¶ 12-13, 5118 ¶ 24 (2015) (“AT&T-Plateau Wireless Order”). See also AT&T-Leap Order, 29 FCC Rcd at 2752-53 ¶¶ 39, 41. For transactions that result in the acquisition of wireless business units and customers or change the number of firms in any market, the Commission also applies an initial screen based on the size of the post-transaction HHI and the change in the HHI. See, e.g., AT&T-Plateau Wireless Order, 30 FCC Rcd at 5118 ¶ 24, n.82. Federal Communications Commission DA 15-1487 39 encounters other factors that may bear on the public interest inquiry.125 In addition, the Commission determined in the Mobile Spectrum Holdings Report and Order that increased aggregation of below-1-GHz spectrum would be treated as an “enhanced factor” under its case-by-case review of license transfers if post-transaction the acquiring entity would hold approximately one-third or more, or 45 megahertz or more, of the currently suitable and available spectrum below 1 GHz.126 54. The Commission includes in its initial screen spectrum that it finds is suitable and available for the provision of mobile telephony/broadband services. Suitability is based upon whether the spectrum band at issue is capable of supporting mobile service given its physical properties and the state of equipment technology, whether the spectrum is licensed with a mobile allocation and corresponding service rules, and whether the spectrum is committed to another use that effectively precludes its uses for the relevant mobile services.127 With respect to availability, the Commission considers particular spectrum to be a relevant input if it is fairly certain that it will meet the criteria for suitability in the near term.128 55. In the past decade, in the context of its review of secondary market transactions, the Commission periodically determined that additional spectrum was suitable and available, and therefore subject to inclusion in the spectrum screen – including 700 MHz,129 Advanced Wireless Service-1 (“AWS-1”),130 Broadband Radio Service (“BRS”),131 and Wireless Communications Service (“WCS”).132 The Commission updated the spectrum screen in the Mobile Spectrum Holdings Report and Order by adding 151 megahertz of spectrum in total from the AWS-4 (2.0/2.2 GHz), H Block (1.9 GHz), and BRS and Educational Broadcast Service (“EBS”) (2.5 GHz) bands.133 It also designated for future inclusion in the spectrum screen, the amount of 600 MHz Band spectrum that would be made available through the incentive auction, and the 65 megahertz of AWS-3 spectrum as it becomes available on a market by market basis.134 The current suitable and available spectrum included in the screen is as follows: 125 See, e.g., Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6221-22 ¶ 225; AT&T-Plateau Wireless Order, 30 FCC Rcd at 5113 ¶ 12. See also AT&T-Leap Order, 29 FCC Rcd at 2752 ¶ 39. 126 See, e.g., Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6240 ¶¶ 282-88. 127 See, e.g., Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6169 ¶ 71; AT&T-Plateau Wireless Order, 30 FCC Rcd at 5116-17 ¶ 21; AT&T WCS Order, 27 FCC Rcd 16459, 16469-70 ¶ 29 n.81. 128 See id. 129 See Applications of AT&T Inc. and Dobson Communications Corporation for Consent To Transfer Control of Licenses and Authorizations, WT Docket No. 07-153, Memorandum Opinion and Order, 22 FCC Rcd 20295, 20307-8 ¶ 17 (2007). 130 See Applications of Sprint Nextel Corporation and Clearwire Corporation for Consent To Transfer Control of Licenses, Leases, and Authorizations, WT Docket No. 08-94, Memorandum Opinion and Order, 23 FCC Rcd 17570, 17599 ¶ 72 (2008) (“Sprint Nextel-Clearwire Order”). 131 Most BRS spectrum is considered available in those markets where the transition of BRS spectrum to the new band plan has been completed. See Sprint Nextel-Clearwire Order, 23 FCC Rcd at 17598-99 ¶ 70; Amendment of Part 27 of the Commission’s Rules to Govern the Operation of Wireless Communications Services in the 2.3 GHz Band, WT Docket No. 07-293, Report and Order, 25 FCC Rcd 11710, 11711 ¶ 1 (2010). 132 See AT&T WCS Order, 27 FCC Rcd at 16470-71 ¶ 31. 133 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6172-87 ¶¶ 82-125. 134 See id., 29 FCC Rcd at 6171-72, 6176-79 ¶¶ 76-81, 94-102. The Commission also subtracted 12.5 megahertz of Specialized Mobile Radio Service (“SMR”) and 10 megahertz that was the Upper 700 MHz D Block. See id., 29 FCC Rcd at 6187-90 ¶¶ 126-34. Federal Communications Commission DA 15-1487 40 Table IV.A.1 Spectrum Included in the Spectrum Screen Spectrum Band Megahertz (Amount) 700 MHz 70 Cellular 50 SMR 14 Broadband PCS 130 AWS-1a 90 H-Block 10 AWS-4 40 WCS 20 BRSb 67.5 EBS 89 Total Amount of Spectrum 580.5 a AWS-1 is not attributable in markets where federal government users have not been relocated. b BRS is not attributable in markets where previous BRS licensees have not been transitioned. 56. For those markets identified by the spectrum screen, or where the Commission encounters other factors that may bear on the public interest inquiry,135 the Commission generally conducts further competitive review to determine whether the proposed transaction would result in an increased incentive or ability for the assignee or transferee to behave in an anticompetitive manner. Further, as well as modifying the spectrum screen, the Mobile Spectrum Holdings Report and Order requires that any increase in spectrum holdings of below 1 GHz be treated as an “enhanced factor” for case-by-case review if post-transaction the acquiring entity would hold approximately one-third or more of the suitable and available spectrum below 1 GHz,136 such that further concentration of such spectrum will not have adverse competitive effects either in particular local markets or on a broader regional or national level.137 The case-by-case review that the Commission undertakes considers factors that are important in predicting the incentives and ability of service providers to successfully reduce competition on price or non-price terms, and evaluates transaction-specific public interest benefits that may mitigate or outweigh any public interest harms that might arise.138 The Commission can condition approval of a transaction 135 For example, the Commission also considered whether harms in numerous local markets may result in nationwide harms and has considered potential harms from concentration in a particular band with an important ecosystem. See, e.g., Applications of Cellco Partnership d/b/a Verizon Wireless and SpectrumCo LLC and Cox TMI, LLC for Consent To Assign AWS-1 Licenses, WT Docket No. 12-4, Memorandum Opinion and Order, 27 FCC Rcd 10698, 10721-22, 10727 ¶¶ 64, 76 (2012) (“Verizon Wireless-SpectrumCo Order”). 136 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6233, 6240 ¶ 267, ¶¶ 286-88. The Commission applied the below-1-GHz review, as set forth in paragraph 286 of the Mobile Spectrum Holdings Report and Order, for the first time in the AT&T-Plateau Wireless Order. More recently, the Commission applied the below-1-GHz review as set forth in paragraph 287 of this Report and Order for the first time in the AT&T-Club 42 Order. See generally AT&T-Plateau Wireless Order, 30 FCC Rcd 5107; Application of AT&T Mobility Spectrum LLC and Club42CM Limited Partnership for Consent To Assign Licenses, WT Docket No. 14-145, Memorandum Opinion and Order, FCC 15-150 (rel. Nov. 12, 2015) (“AT&T- Club 42 Order”). 137 See Mobile Spectrum Holdings Report and Order, 29 FCC Rcd at 6240 ¶ 288. 138 These competitive variables include, but are not limited to: the total number of rival service providers; the number of rival firms that can offer competitive service plans; the coverage by technology of the firms’ respective networks; the rival firms’ Federal Communications Commission DA 15-1487 41 on the divestiture of licenses or certain other commitments in markets where necessary to find an application serves the public interest.139 57. Since the Seventeenth Report, a number of small transactions involving the transfer of spectrum licenses, as well as, in certain cases, network infrastructure and other assets, have been filed with the Commission.140 Among the smaller transactions that have occurred in the past couple of years are a number in which a nationwide service provider acquired spectrum or other assets from a small or regional licensee. Not including intra-market spectrum swaps of equal amounts of spectrum or transactions involving increased aggregation of low-band spectrum, from January 2014 through June 2015 the Commission approved approximately 110 applications in total filed by the four nationwide providers to acquire PCS, AWS-1, Cellular, and/or 700 MHz licenses from a non-nationwide licensee or lease additional BRS/EBS spectrum from a non- nationwide licensee. c. Additional Spectrum Initiatives 58. The 3.5 GHz Band141 proceeding is a Commission initiative that will make more spectrum available to facilitate the provision of mobile wireless service. On April 17, 2015, the Commission adopted a Report and Order that will advance the use of this band.142 This Report and Order established a three-tiered spectrum authorization framework to facilitate a variety of small cell and other broadband uses of the 3.5 GHz Band143 on a shared basis with incumbent federal and non-federal users of the band.144 The three tiers of users, in order of priority, are: Incumbent Access, Priority Access, and General Authorized Access (“GAA”).145 Incumbent Access users include: (1) military radar systems; (2) non-federal fixed satellite service (“FSS”) earth stations; and (3) for a finite period, grandfathered terrestrial wireless broadband service licensees in the 3650- market shares; the combined entity’s post-transaction market share and how that share changes as a result of the transaction; the amount of spectrum suitable for the provision of mobile telephony/broadband services controlled by the combined entity; and the spectrum holdings of each of the rival service providers. See, e.g., AT&T-Plateau Wireless Order, 30 FCC Rcd at 5120 ¶ 29; AT&T-Leap Order, 29 FCC Rcd at 2745-46 ¶ 21. 139 See, e.g., AT&T-Leap Order, 29 FCC Rcd at 2743-44 ¶ 16; Verizon Wireless-SpectrumCo Order, 27 FCC Rcd at 10711 ¶ 30. 140 See, e.g., AT&T-Club 42 Order, FCC 15-150; AT&T-Plateau Wireless Order, 30 FCC Rcd 5107; Application of AT&T Mobility Spectrum LLC and Kaplan Telephone Company, Inc. for Consent To Assign Licenses, WT Docket 14-167, Memorandum Opinion and Order, 30 FCC Rcd 8502 (WTB 2015); Applications of AT&T Mobility Spectrum LLC and KanOkla Telephone Association, WT Docket No. 14-199, Memorandum Opinion and Order, 30 FCC Rcd 8555 (WTB 2015); Application of AT&T Mobility Puerto Rico Inc. and Worldcall Inc. for Consent To Assign Licenses, WT Docket No. 14-206, Memorandum Opinion and Order, 30 FCC Rcd 9763 (WTB 2015); Application of AT&T Mobility Spectrum LLC and Consolidated Telephone Company for Consent To Assign Licenses, WT Docket No. 14-254, Memorandum Opinion and Order, 30 FCC Rcd 9797 (WTB 2015); Application of Hardy Cellular Telephone Company and McBride Spectrum Partners, LLC for Consent To Assign License, WT Docket No. 14-240, Memorandum Opinion and Order, 30 FCC Rcd 9899 (WTB 2015); Application of TeleGuam Holdings, LLC and Club 42 CM Limited Partnership for Consent To Assign Licenses, WT Docket No. 15-7, Memorandum Opinion and Order, 30 FCC Rcd 10213 (WTB 2015). 141 The 3.5 GHz Band encompasses 3550-3700 MHz. See Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Report and Order and Second Further Notice of Proposed Rulemaking, 30 FCC Rcd 3959, 3961 ¶ 1 (2015) (“3.5 GHz Order and 2nd FNPRM”). 142 See generally 3.5 GHz Order and 2nd FNPRM, 30 FCC Rcd 3959. 143 Priority Access Licenses will not be available in the 3650-3700 MHz portion of the band, which is reserved for GAA and Grandfathered Wireless Broadband Licensees. See id., 30 FCC Rcd at 3978 ¶ 54. 144 See id., 30 FCC Rcd at 3967 ¶ 24, 3978 ¶ 54. 145 See id. Federal Communications Commission DA 15-1487 42 3700 MHz portion of the band.146 These users will be protected from harmful interference from Priority Access and GAA users. Priority Access licensees147 will receive protection from interference from GAA users.148 Access and operations will be coordinated by a dynamic spectrum access system (“SAS”),149 conceptually similar to – but more technologically advanced than – the databases used to manage Television White Spaces devices.150 The innovative spectrum sharing techniques adopted in this Order will allow the introduction of 150 megahertz of contiguous spectrum for the Citizens Broadband Radio Service, while protecting critical federal uses, which will enable the exploration of new technologies and spectrum sharing with a focus on relatively low-powered applications. 59. The Commission has also taken steps to explore other technologies,151 including, for example, uses of spectrum above 24 GHz. On October 22, 2015, the Commission adopted a Notice of Proposed Rulemaking (“NPRM”) promoting uses of spectrum above 24 GHz for commercial purposes.152 The Commission’s “spectrum frontiers” proceeding holds the potential to unlock vast millimeter-wave bands for mobile use, particularly for use by Fifth Generation (“5G”) mobile services.153 Previously, bands above 24 GHz were believed to be infeasible for mobile use due to their straight line propagation and atmospheric absorption characteristics.154 However, as technologies continue to evolve, innovators are working to tap into the potential of using millimeter-wave bands for mobile services,155 and the promise of high capacity data transfers from these millimeter-wave bands could be a useful supplement to the mobile services offered in lower bands. The NPRM also proposes a variety of licensing mechanisms with the goal of developing flexible rules that will accommodate a wide variety of current and future technologies.156 146 See id., 30 FCC Rcd at 3961-62 ¶¶ 3-4, 4035-40 ¶¶ 247-68, 4042-48 ¶¶ 276-96; see also Wireless Telecommunications Bureau Seeks Comment On An Appropriate Method for Determining The Protected Contours For Grandfathered 3650-3700 MHz Band Licensees, GN Docket No. 12-354, Public Notice, DA 15-1208 (WTB rel. Oct. 23, 2015). 147 A Priority Access License (“PAL”) is defined as a non-renewable authorization to use a 10 megahertz channel in a single census tract for three years. PALs will be assigned via competitive bidding in up to seventy megahertz of the 3550-3650 MHz portion of the band. See 3.5 GHz Order and 2nd FNPRM, 30 FCC Rcd at 3961 ¶ 4. One licensee may hold up to forty megahertz of PALs in any given census tract at any given time. See id., 30 FCC Rcd at 3998 ¶ 117. 148 GAA users could be a diverse group of stakeholders, including consumers, enterprises, and service providers. See 3.5 GHz Order and 2nd FNPRM, 30 FCC Rcd at 4009 ¶ 156. GAA users would be permitted opportunistic use of all spectrum from 3550-3700 MHz that is not being used at the time by PAL holders or incumbents. See id., 30 FCC Rcd at 3983 ¶¶ 72- 73. The definition of “use” has not yet been resolved. See id., 30 FCC Rcd at 4081 ¶ 419. 149 See id., 30 FCC Rcd at 3962 ¶ 4, 3984-87 ¶¶ 75-86, section III.H. 150 See id., 30 FCC Rcd at 4035 ¶ 247, 4069-71 ¶¶ 379-86. 151 We note that in May, 2015, the Commission released a Public Notice seeking more information on Long Term Evolution (“LTE”)-Unlicensed. The Commission plans to continue to monitor the development of unlicensed technologies, including LTE-U. See Office of Engineering and Technology and Wireless Telecommunications Bureau Seek Information on Current Trends in LTE-U and LAA Technology, ET Docket No. 15-105, Public Notice, 30 FCC Rcd 4457 (2015) (“LTE-U Public Notice”). 152 See Use of Spectrum Bands Above 24 GHz for Mobile Radio Services, GN Docket No. 14-177, Notice of Proposed Rulemaking, FCC 15-138 (rel. Oct. 23, 2015) (“Spectrum Frontiers NPRM”). 153 See id., FCC 15-138, at ¶¶ 1, 6. We note that we do not intend to define what qualifies as “5G.” Standard bodies like 3GPP and the International Telecommunications Union (“ITU”) plan to develop the requirements by early 2017. See “Tentative 3GPP Timeline for 5G,” available at http://www.3gpp.org/news-events/3gpp-news/1674-timeline_5g (Mar. 2015). See also id., FCC 15-138, at ¶ 1. 154 See id., FCC 15-138, at ¶ 5. 155 See id., FCC 15-138, at ¶¶ 5, 12. 156 See id., FCC 15-138, at ¶ 3. Federal Communications Commission DA 15-1487 43 3. Analysis of Spectrum Holdings 60. Table IV.A.2 (Percentage Spectrum Holdings, by Provider, by Frequency Band) and Table IV.A.3 (Population-Weighted Average Megahertz Holdings by Provider, by Frequency Band) below present spectrum holdings by service provider including all spectrum bands currently considered suitable and available. Table IV.A.3 shows megahertz holdings for each service provider, weighted by population, and Chart IV.A.1 is a graph of providers’ spectrum holdings by frequency band, measured on a MHz-POPs basis. As of October 2015, Verizon Wireless, AT&T, Sprint, and T-Mobile, together, hold over 80 percent of all spectrum suitable and available for the provision of mobile wireless services, measured on a MHz-POPs basis. Table IV.A.2 Percentage Spectrum Holdings, Measured on a MHz-POPs Basis by Licensee, by Frequency Band* 700 MHz Cell. SMR PCS H Block AWS- 1 AWS- 4 WCS BRS EBS Spectrum 70 MHz 50 MHz 14 MHz 130 MHz 10 MHz 90 MHz 40 MHz 20 MHz 67.5 MHz 112.5 MHz*** VZW 31.0% 48.1% 0.0% 16.3% 0.0% 38.8% 0.0% 0.0% 0.0% 0.0% AT&T 41.6% 44.6% 0.0% 29.4% 0.0% 16.3% 0.0% 100.0% 0.0% 0.0% Sprint 0.0% 0.0% 96.5% 28.0% 0.0% 0.0% 0.0% 0.0% 86.8% 69.8% T-Mobile 9.5% 0.1% 0.0% 21.8% 0.0% 40.9% 0.0% 0.0% 0.0% 0.0% USCC 3.6% 4.2% 0.0% 1.3% 0.0% 0.6% 0.0% 0.0% 0.0% 0.0% DISH** 6.6% 0.0% 0.0% 0.0% 100.0% 0.0% 100.0% 0.0% 0.0% 0.0% Other*** 7.8% 3.0% 3.5% 3.1% 0.0% 3.4% 0.0% 0.0% 13.2% 30.2% * Staff estimates as of Oct. 19, 2015. Abbreviations for spectrum bands: Cell. (Cellular); SMR (Specialized Mobile Radio Service), BRS (Broadband Radio Service), EBS (Educational Broadband Service). ** Dish Network Corporation currently does not provide mobile service. *** In the application of the spectrum screen in secondary market transactions, 89 megahertz of EBS spectrum is included. Table IV.A.3 Population-Weighted Average Megahertz Holdings by Licensee, by Frequency Band* 700 MHz Cell. SMR PCS H Block AWS -1 AWS -4 WCS BRS EBS Spectrum Counted 70 MHz 50 MHz 14 MHz 130 MHz 10 MHz 90 MHz 40 MHz 20 MHz 67.5 MHz 112.5 MHz*** VZW 21.7 24.6 0.0 21.3 0.0 34.9 0.0 0.0 0.0 0.0 AT&T 29.1 22.9 0.0 38.4 0.0 14.7 0.0 20.0 0.0 0.0 Sprint 0.0 0.0 13.9 36.6 0.0 0.0 0.0 0.0 58.6 78.5 T-Mobile 6.7 0.0 0.0 28.4 0.0 36.8 0.0 0.0 0.0 0.0 USCC 2.5 2.2 0.0 1.7 0.0 0.6 0.0 0.0 0.0 0.0 DISH** 4.6 0.0 0.0 0.0 10.0 0.0 40.0 0.0 0.0 0.0 Other*** 5.4 1.5 0.5 4.1 0.0 3.0 0.0 0.0 8.9 34.0 * Staff estimates as of Oct. 19, 2015. ** Dish Network Corporation currently does not provide mobile service. *** In the application of the spectrum screen in secondary market transactions, 89 megahertz of EBS spectrum is included. Federal Communications Commission DA 15-1487 44 Note: Staff estimates as of Oct. 19, 2015. 61. Chart IV.A.1 above shows the population-weighted spectrum holdings of nationwide wireless service providers by frequency. It provides a side-by-side comparison of each licensee’s total spectrum holdings by band, measured by population-weighted average megahertz.157 All four nationwide service providers hold substantial amounts of above-1-GHz spectrum. Verizon Wireless, AT&T, and T-Mobile each hold a substantial number of PCS and AWS-1 spectrum licenses, while Sprint holds significant amounts of PCS spectrum. Verizon Wireless holds approximately 26 percent of the licensed MHz-POPs of the combined PCS and AWS-1 band spectrum, while the comparable percentages are approximately 24 percent for AT&T, approximately 17 percent for Sprint, and approximately 30 percent for T-Mobile. Regional service provider, US Cellular, holds approximately one percent of the combined PCS and AWS-1 band spectrum, while other smaller service providers hold the remainder. In addition to its PCS and AWS-1 holdings, AT&T holds all 20 megahertz of the licensed WCS spectrum,158 while Sprint holds a predominant amount of 2.5 GHz spectrum, comprised of the BRS and EBS bands, the highest frequencies currently considered “suitable” and “available” for the provision of mobile broadband service.159 Finally, while the granted AWS-3 spectrum licenses are not currently included in the screen, we note that AT&T holds approximately 36 percent of the licensed MHz-POPs, while Verizon Wireless, T-Mobile, and DISH each hold approximately 20 percent, 6 percent, and 37 percent, respectively, and other smaller service providers hold the remaining 1 percent. 157 We consider population-weighted spectrum holdings in order to account for customer density in different geographic areas. A spectrum license in Los Angeles or New York City, for example, covers more customers than a spectrum license over the same amount of land area in White Sands, New Mexico. 158 See AT&T WCS Order, 27 FCC Rcd at 16461-62 ¶¶ 4-6; Application of AT&T and Sprint Seek FCC Consent to the Assignment of WCS Licenses, WT Docket No. 14-83, ULS File No. 0006344543 (filed May 6, 2014); Application, Appendix A – Spectrum Aggregation. See also Application of AT&T and Sprint Seek FCC Consent to the Assignment of WCS Licenses, WT Docket No. 14-83, Public Notice, 29 FCC Rcd 5879, 5879-80 (WTB 2014). 159 See SoftBank-Sprint-Clearwire Order, 28 FCC Rcd at 9645 ¶ 11. - 10 20 30 40 50 60 70 Verizon Wireless AT&T Sprint T-Mobile US Cellular DISH Other M H z- P O P s (b il li o n s) Chart IV.A.1 Mobile Wireless Provider Spectrum Holdings by Band Weighted by Population WCS EBS BRS AWS-4 AWS-1 H Block PCS SMR Cellular 700 MHz Federal Communications Commission DA 15-1487 45 62. Below-1-GHz spectrum currently includes the 700 MHz band, SMR (800 MHz), and Cellular (850 MHz) spectrum. The two largest nationwide service providers, AT&T and Verizon Wireless, each hold a significant amount of the available low-band spectrum. In particular, when measured on a licensed MHz-POP basis, AT&T holds approximately 38 percent, while Verizon Wireless holds approximately 35 percent. In addition, Sprint holds approximately 10 percent, T-Mobile holds approximately five percent, and a number of other smaller licensees, combined, hold the remaining approximately 12 percent. Service providers also vary with respect to their below-1-GHz spectrum holdings according to population density, as seen in Chart IV.A.2 below. Specifically, AT&T and T-Mobile hold relatively more of their low-band spectrum in urban areas, Sprint’s and Verizon Wireless’s low-band spectrum covers both urban and rural areas, and the other smaller licensees hold more low-band spectrum in rural areas than in urban areas. Chart IV.A.2 Average Below-1-GHz Spectrum by Population Density Deciles, Oct. 2015 B. Non-Spectrum Input Segments 1. Wireless Infrastructure 63. Wireless infrastructure facilities are one of the major inputs in the provision of mobile wireless services and host cellular base stations.160 In addition to the use of towers and other tall structures, wireless 160 These facilities include towers and other tall structures for macro sites, such as lattice towers, guyed towers, monopoles, rooftops, water towers, and steeples, for macro sites. Verizon Wireless AT&T Sprint Nextel T-Mobile US Cellular 0 5 10 15 20 25 30 35 40 45 50 55 60 65 1st Decile 2nd Decile 3rd Decile 4th Decile 5th Decile 6th Decile 7th Decile 8th Decile 9th Decile 10th Decile 0.1% 0.4% 0.5% 0.7% 1.3% 2.1% 3.1% 5.8% 11.5% 74.4% A V E R A G E S U B 1 G H Z S P E C T R U M ( P O P - W E I G H T E D M H Z ) US POPULATION DECILES BY COUNTY POPULATION DENSITY % of Pops % of Area More Dense Less Dense Other DISH Federal Communications Commission DA 15-1487 46 infrastructure also includes distributed antenna systems (“DAS”)161 and facilities for small cell technologies162 that are generally deployed to address coverage and capacity issues indoors, in densely populated areas outdoors, and even underground.163 For example, small cells and DAS antennas can be placed on utility poles, buildings, or traffic signal poles, in areas where constructing towers is not feasible or wireless traffic demands are too great to be met solely with fewer large cells.164 64. In order to expand geographic service area coverage, to improve coverage in existing service areas, and to accommodate newer technologies, mobile service providers have historically deployed more cell sites. However, after many years of consecutive growth, the number of cell sites in use by mobile service providers appears to have stabilized. According to CTIA, there were 298,055 cell sites in use at year-end 2014, down approximately 2 percent (or 6,305) from 304,360 as of year-end 2013,165 after a continuous increase since 1985.166 In addition to macro cell sites, mobile service providers, in recent years, have started to deploy small cells and DAS sites to fill local coverage gaps or to increase local capacity.167 65. A specialized communications tower industry has developed to provide and manage the support structures for the cell sites, and leases space to mobile wireless service providers. Today, there are more than 110 tower and DAS operators in the United States,168 and a majority of towers are now owned or operated by 161 A DAS network consists of three primary components: (i) a number of remote communications nodes (DAS node(s)), each including at least one antenna for the transmission and reception of a wireless service provider’s RF signals; (ii) a high capacity signal transport medium (typically fiber optic cable) connecting each DAS node back to a central communications hub site; and (iii) radio transceivers or other head-end equipment located at the hub site that propagates and/or converts, processes or controls the communications signals transmitted and received through the DAS nodes. See The DAS Forum, “Distributed Antenna Systems (DAS) and Small Cell Technologies Distinguished,” at 3 (Feb. 2013), available at http://www.thedasforum.org/resources/send/2-resources/24-das-and-small-cell-technologies-distinguished. 162 “Small cells” is an umbrella term for operator-controlled, low-powered radio access nodes, including those that operate in licensed spectrum and unlicensed carrier-grade Wi-Fi. Small cells typically have a range from 10 meters (e.g., femtocells) to several hundred meters (e.g., microcells). See Small Cell Forum, “Small Cell Definition,” available at http://www.smallcellforum.org/about/about-small-cells/small-cell-definition/. 163 See The DAS Forum, “Distributed Antenna Systems (DAS) and Small Cell Technologies Distinguished,” at 6 (Feb. 2013), available at http://www.thedasforum.org/resources/send/2-resources/24-das-and-small-cell-technologies-distinguished. 164 Because DAS sites are less visible than tower structures, they may be particularly desirable in areas with stringent siting regulations, such as historic districts. 165 See CTIA Year-End 2014 Wireless Indices Report, at p. 101. Because multiple cell sites can be co-located in the same “tower” site, the reported cell sites should not be equated with “towers.” The reported cell sites include repeaters and other cell-extending devices (e.g., femtocells, or distributed antenna systems). See id. at pp. 101, 102. 166 See id., at p. 104. The decrease in the total number of commercial cell sites in 2014 is likely due to “a combination of consolidation and the retirement of older generation of technologies.” Id at p. 101. 167 See Phil Goldstein, “AT&T drops goal of deploying 40,000 small cells by end of 2015, citing benefits of Leap deal” (Mar. 5, 2015), available at http://www.fiercewireless.com/story/att-drops-goal-deploying-40000-small-cells-end-2015-citing- benefits-leap-de/2015-03-05. See also Jeff Moore, “Verizon facing small cell 'deployment difficulties,' but forging ahead aggressively” (July 29, 2015), available at http://www.fierceinstaller.com/story/verizon-facing-small-cell-deployment- difficulties-forging-ahead-aggressivel/2015-07-29; “T-Mobile and Nokia aim small: to deploy small LTE cells that use unlicensed spectrum” (Feb. 24, 2015), available at http://www.phonearena.com/news/T-Mobile-and-Nokia-aim-small-to- deploy-small-LTE-cells-that-use-unlicensed-spectrum_id66327; “Sprint to add ‘tens of thousands’ of small cells, bring 800 MHz and 2.5 GHz LTE to ‘nearly all’ macro sites” (Aug. 4, 2015), available at http://www.fiercewireless.com/story/sprint- add-tens-thousands-small-cells-bring-800-mhz-and-25-ghz-lte-nearly-a/2015-08-04. 168 See Wireless Estimator, “Top 100 Tower Companies in the U.S.,” available at http://wirelessestimator.com/top-100-us- tower-companies-list/. Federal Communications Commission DA 15-1487 47 independent companies rather than by mobile wireless service providers.169 Independent tower operators own, operate and lease shared wireless communications and broadcasting towers, manage other tall structure sites (such as rooftops, and water towers), and to a lesser extent, build and operate DAS networks and small cell facilities for mobile service providers.170 In most cases, tower operators and property owners lease antenna, rooftop and other site space to multiple wireless service providers.171 One estimate indicates that the three largest publicly traded neutral host providers (Crown Castle, American Tower, and SBA Communications) own or operate more than 94,540 towers as of August 2015.172 The availability of leased space on existing towers for mobile wireless service providers may eliminate the need to build new towers, reduce the capital requirements for network deployment and capacity expansion, and facilitate the entry of new wireless service providers. As of July 2015, Chart IV.B.1 shows that there were three or more tower operators in 90 percent of counties, and six or more in 41 percent of counties based on data collected from 34 tower providers173 in the United States in June and July 2015.174 169 Some major wireless service providers have sold or in the process of selling their tower business to third party tower operators. See American Tower News Release, “American Tower Corporation Announces Verizon Tower Portfolio Transaction” (Feb. 5, 2015); AT&T News Release, “AT&T and Crown Castle Close $4.83 Billion Tower Transaction” (Dec. 16, 2013); Crown Castle News Release, “Crown Castle Completes Tower Transaction With T-Mobile USA” (Nov. 30, 2012); Sprint Nextel News Release, “Sprint Nextel Completes Tower Sale to TowerCo for Approximately $670 Million in Cash” (Sept. 24, 2008). 170 See American Tower 2014 Annual Report (10-K) at 1, available at http://www.americantower.com/corporateus/investor- relations/annual-reports-proxy-statements/index.htm. Crown Castle 2014 Annual Report (10-K) at 1, available at http://investor.crowncastle.com/phoenix.zhtml?c=107530&p=irol-sec&seccat01.1_rs=49&seccat01.1_rc=6. SBA Communications 2014 Annual Report (10-K) at 1, available at http://ir.sbasite.com/reports.cfm. 171 See American Tower, 2014 Annual Report at 1 (“Our primary business is leasing antenna space on multiple-tenant communications sites to wireless service providers, radio and television broadcast companies, wireless data providers, government agencies and municipalities and tenants in a number of other industries”). See also Verizon Network Real Estate Inquires (“Verizon Wireless receives thousands of inquiries each year from property owners, property managers and customers who offer property on which our communications facilities can be located”), available at http://www.verizonwireless.com/b2c/realestate/. 172 See Wireless Estimator, “Top 100 Tower Companies in the U.S.,” available at http://www.wirelessestimator.com/t_content.cfm?pagename=US-Cell-Tower-Companies-Complete-List (list of tower operators, with Crown Castle at 39,928, American Tower at 39,739, and SBA Communications at 14,873 as of August 2015, not including DAS and small cells). 173 Tower site information was either downloaded from the tower provider’s website or requested via telephone in June and July 2015. See Wireless Estimator, “Top 100 Tower Companies in the U.S.,” (many tower providers’ websites), available at http://www.wirelessestimator.com/t_content.cfm?pagename=US-Cell-Tower-Companies-Complete-List. The 34 tower providers listed in this Report are Airwave Strategies, American Tower Corporation, AT&T, Badger Towers, Crown Castle, Communication Enhancement, Clearview Tower Company, Central States Tower Holdings, Com Sites West, CTI Towers, Diamond Communications, Day Wireless Systems, ERS Antenna Site Management, Hemphill Tower, Horizon Tower, Horvath Communications, InSite Towers, Industrial Tower and Wireless, KGI Wireless, Message Center Management, Nsight Tower Holdings, Pegasus Tower Company, SBA Communications Corporation, Skyway Towers, Sprint, Subcarrier Communications, Tarpon Towers, T-Mobile, Tower Acquisition, TowerCo, Tower Ventures, Unison Site Management, Vertical Bridge Holdings, and Wireless Properties. 174 Excluding Alaska, Guam, Puerto Rico, and the U.S. Virgin Islands. In the Seventeenth Report, 50% of counties had more than six tower operators based on data collected from 11 tower providers in September 2013. See Seventeenth Report, 29 FCC Rcd at 15367 ¶ 111. Given that we analyzed data from 34 tower providers in this Report, the data are not comparable across the two Reports, although we note that the decrease from 50% to 40% in this Report, despite the fact that more tower operators’ data are included, is likely due to a combination of consolidation and the retirement of old technologies. See CTIA Year-End 2014 Wireless Indices Report, at p. 101. For example, American Tower Corporation (“ATC”) bought General Tower Partners (“GTP”) in September 2013. Crown Castle bought AT&T Towers in December 2013. Federal Communications Commission DA 15-1487 48 Source: Data collected from 34 tower companies (data include towers and rooftops) in June and July 2015. 66. As shown in Chart IV.B.2 below, tower operators tend to build and operate more towers in more densely populated areas. For example, as of July 2015, the average number of tower sites per county is 33 for counties with a population density between 75 and 100 people per square mile, compared to 300 per county for counties with a population density between 2000 and 4000 people per square mile. Note: Data based on 34 tower companies referenced above. Counties considered rural are those with fewer than 100 people per square mile. Population density is from the 2010 U.S. Census. 96 228 362 553 595 466 290 180 135 91 47 23 17 13 0 100 200 300 400 500 600 700 1 2 3 4 5 6 7 8 9 10 11 12 13 14+ N u m b e r o f C o u n ti e s Number of Site Operators in a County Chart IV.B.1 Number of Counties with a Certain Number of Site Operators 5.0 9.3 15.4 20.9 27.9 33.3 50.8 55.0 85.6 103.4 140.8 139.0 210.8 299.9 191.5 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 <=1 (1, 10] (10, 25] (25, 50] (50, 75] (75, 100] (100, 150] (150, 200] (200, 250] (250, 500] (500, 750] (750, 1000] (1000, 2000] (2000, 4000] >4000 A v e S it e C o u n ts p er C o u n ty County Population Density Range (persons per square mile) Chart IV.B.2 Average Tower Site Counts per County vs. Pop Density 2015 Rural Federal Communications Commission DA 15-1487 49 67. In addition, as shown in Chart IV.B.3.a below, there are more tower operators in densely populated counties than in less densely populated counties. The number of tower site operators per county ranges from two site operators per county in the least densely populated counties to more than seven site operators in the most densely populated counties. Chart IV.B.3.b below indicates that counties with larger land areas generally have more tower operators than counties with smaller land areas. Source: Data are based on the 34 tower companies referenced above. Rural counties are those with fewer than 100 people per square mile. Population density and land area are from the 2010 U.S. Census. Source: Data are based on the 34 tower companies referenced above. Rural counties are those with fewer than 100 people per square mile. Population density and land area are from the 2010 U.S. Census. 2.2 3.2 4.1 4.8 5.5 5.7 6.4 6.5 7.1 7.8 8.6 8.0 9.3 9.2 8.5 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 <=1 (1, 10] (10, 25] (25, 50] (50, 75] (75, 100] (100, 150] (150, 200] (200, 250] (250, 500] (500, 750] (750, 1000] (1000, 2000] (2000, 4000] >4000 A v e. T o w er O p er a to r C o u n ts p er C o u n ty County Population Density Range (persons per square mile) Chart IV.B.3.a Ave. Tower Operator Counts per County vs. Pop Density 2015 2.6 4.7 6.5 8.3 5.4 5.4 4.8 4.7 5.3 5.4 5.4 5.1 5.3 6.4 7.0 6.4 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 A v e T o w er O p er a to r C o u n ts p er C o u n ty County Land Area Range (square mile) Chart IV.B.3.b Ave. Tower Operator Counts per County vs. County Land Area 2015 Federal Communications Commission DA 15-1487 50 68. There are two factors that have a significant effect on the deployment or modification of tower and DAS sites: (i) capital expenditure; and (ii) obtaining the necessary regulatory and zoning approvals from local and federal authorities.175 In terms of capital expenditure, co-locating wireless equipment on existing structures is often the most efficient and economical solution for mobile wireless service providers that need new cell sites, either to expand their existing coverage area, increase their capacity, or deploy a new generation of mobile broadband technology. The average cost to build a new tower is between $250,000 and $300,000, whereas the average cost of co-location on an existing tower is about 25 percent of the total cost of a new tower.176 The three largest publicly-traded infrastructure companies alone made capital expenditures of approximately $1.58 billion in building new sites and upgrading existing sites in 2014, up from $1.17 billion in 2013.177 The total annual expenditure for structures by wireless service providers (excluding satellite service providers) was $9.952 billion for 2013, almost a 74 percent increase from $5.723 billion in 2012.178 2. Backhaul 69. Backhaul facilities link a mobile wireless service provider’s cell sites to the mobile switching centers that provide connections to the provider’s core network, the public switched telephone network, or the Internet, carrying wireless voice and data traffic for routing and onward transmission. Backhaul connections are an integral component of a wireless service provider’s network, and the cost of backhaul is approximately 30 percent of the operating cost of providing wireless service.179 Backhaul services are generally provided by incumbent local exchange carriers (“ILECs”), competitive local exchange carriers (“CLECs”), competitive fiber and microwave wholesalers, cable providers, and independent backhaul operators.180 70. As mobile data traffic has grown rapidly in recent years, the leading mobile wireless service providers have deployed or are in the process of deploying Ethernet backhaul either over fiber or microwave to their cell sites. For example, as of June 2013, over 90 percent of AT&T’s data traffic was already on enhanced backhaul.181 As of March 2015, of its 54,000 cell sites, T-Mobile already has fiber backhaul connections to 50,000 sites.182 Sprint’s network modernization was substantially completed in 2014 which utilizes Ethernet for 175 See Seventeen Report, 29 FCC Rcd at 15369-70 ¶ 114. See also IV.B.3. infra (recent FCC initiatives on wireless facilities siting). 176 See American Tower, 2015 Q1 Earnings Presentation at 12 ($850 - $950 million to build 2,750 - 3,250 new towers, or $292K - $309K per new tower), available at http://www.americantower.com/corporateus/investor-relations/earnings- materials/index.htm. See also Statistic Brain Research Institute, “Cell Phone Tower Statistics” (average cost of building a cell phone tower is $145K, while the average yearly cell phone tower lease rate is $45K, about 25% of the cost of building a new cell tower) (Mar. 15, 2015), available at http://www.statisticbrain.com/cell-phone-tower-statistics/. 177 See PCIA Comments at 6, 7. 178 See CTIA Year-End 2014 Wireless Indices Report, Chart 20, at p. 108 (citing information from the U.S. Census and CTIA). 179 See Sprint Comments at 7. 180 See FierceTelecom.com, “Telco Backhaul Strategies,” at 1-2 (November 2011). Providers of backhaul services include ILECs such as AT&T, Verizon, and CenturyLink; CLECs such as Level 3, tw telecom inc., Cbeyond, Inc., and XO Communications; competitive fiber and microwave wholesalers such as Level 3, FPL FiberNet, IP Networks, and Zayo; cable providers such as Charter Communications, Comcast Business, Cox Carrier Services, and Time Warner Cable Business Class; independent backhaul operators, including backhaul specialists such as Telecom Transport Management, and Tower Cloud, and potentially certain tower operators. 181 See “AT&T 4G LTE Network Ranked Fastest For Second Straight Year,” AT&T news release (June 10, 2013). 182 See Phil Goldstein, “T-Mobile’s Carter: We’d be a ‘very interesting’ partner for Dish” (Mar. 6, 2015), available at http://www.fiercewireless.com/story/t-mobiles-carter-wed-be-very-interesting-partner-dish/2015-03-05. Federal Communications Commission DA 15-1487 51 its backhaul183 and its LTE network covered more than 280 million people as of May 2015.184 Some analysts have predicted that Sprint will use some of its 2.5 GHz spectrum as an alternative to fiber backhaul.185 Verizon Wireless also deployed fiber backhaul facilities for its 4G LTE sites,186 and its LTE network carried 87 percent of its data traffic as of the second quarter of 2015.187 3. Recent FCC Initiatives 71. Section 6409(a) of the Middle Class Tax Relief and Job Creation Act of 2012 provides that a state or local government “may not deny, and shall approve” any request for co-location, removal, or replacement of transmission equipment on an existing wireless tower or base station, provided this action does not substantially change the physical dimensions of the tower or base station.188 On October 21, 2014, the Commission released an order that implemented section 6409(a) and took other actions to eliminate unnecessary reviews, and therefore reduce costs and delays, for wireless facilities siting.189 The Commission has also entered into two Nationwide Programmatic Agreements (“NPAs”) with the Advisory Council on Historic Preservation (“ACHP”) and the National Conference of State Historic Preservation Officers (“NCSHPOs”) to clarify the National Historic Preservation Act (“NHPA”) section 106 process for (1) new tower construction, and (2) collocations of communications equipment on existing towers and other structures.190 In August 2014, the Commission adopted a Report and Order to streamline and eliminate outdated provisions of the Part 17 Rules governing the construction, marking, and lighting of antenna structures.191 In addition, the Commission has taken steps with relevant government and non-governmental stakeholders to develop a process for “clearing” existing towers that did not complete section 106 historic preservation review prior to construction, including “twilight towers” that were constructed prior to the specification of detailed section 106 review procedures in the 2005 NHPA,192 while respecting historic preservation values. This effort will potentially make thousands of additional towers available 183 See Sprint’s 10-K filing (Annual Report) for 2014, at 27, available at http://investors.sprint.com/financial-information/sec- filings/sec-filings-details/default.aspx?FilingId=10724268. 184 See Sprint Comments at 5. 185 See e.g., Phil Goldstein, “Analysts: Sprint’s Network Densification Plans Up in the Air, But Wireless Backhaul Could Play a Key Role” (July 17, 2015), available at http://www.fiercewireless.com/story/analysts-sprints-network-densification- plans-air-wireless-backhaul-could-pl/2015-07-17. 186 See Transcript for Verizon at Oppenheimer Holdings Inc Technology, Internet & Communications Conference, at 7 (Aug. 15, 2012), available at http://www.verizon.com/about/investors/oppenheimers-15th-annual-technology-internet- communications-conference. 187 See Verizon presentation for the 2nd Quarter 2015 earnings results, at 8, available at http://www.verizon.com/about/investors/quarterly-reports/2q-2015-quarter-earnings-conference-call-webcast. 188 Middle Class Tax Relief and Job Creation Act of 2012, Pub. L. No. 112-96, § 6409(a) (2012), codified at 47 U.S.C. § 1455(a). 189 See Acceleration of Broadband Deployment by Improving Wireless Facilities Siting Policies; Acceleration of Broadband Deployment: Expanding the Reach and Reducing the Cost of Broadband Deployment by Improving Policies Regarding Public Rights of Way and Wireless Facilities Siting; 2012 Biennial Review of Telecommunications Regulations, WT Docket Nos. 13-238, WT Docket No. 11-59, WT Docket No. 13-32, Report and Order, 29 FCC Rcd 12865 (2014) (“Wireless Infrastructure Report and Order”). 190 47 C.F.R. Pt. 1, Apps. B and C; see Section 106 of the National Historic Preservation Act, 54 U.S.C. § 306108. 191 See 2004 and 2006 Biennial Regulatory Reviews – Streamlining and Other Revisions of Parts 1 and 17 of the Commission’s Rules Governing Construction, Marking and Lighting of Antenna Structures, WT Docket No. 10-88, Report and Order, 29 FCC Rcd 9787 (2014). 192 Specifically, “twilight towers” are non-compliant towers built between March 16, 2001 and March 7, 2005. Federal Communications Commission DA 15-1487 52 for co-location.193 The Commission has also initiated a process to develop further exclusions from section 106 review for DAS and small cell facilities that are unlikely to adversely affect historic properties.194 72. The Commission has also examined issues related to backhaul including special access services and the use of microwave spectrum for backhaul services in the past few years.195 On September 15, 2014 the Commission’s Wireline Competition Bureau issued an Order on Reconsideration that set a deadline of December 15, 2014 for service providers to submit data on networks, prices, and terms for special access in order to enable the Commission to assess the state of competition.196 Although the Bureau kept the December 15, 2014 deadline for streamlined certifications, it extended this deadline to January 29, 2015 for large businesses with more than 1500 employees, and to February 27, 2015 for small businesses with less than 1500 employees.197 Comments on the data submissions are currently due on January 6, 2016, with replies due on February 6, 2016.198 In addition, on October 16, 2015, the Wireline Competition Bureau initiated an investigation into the terms and conditions of certain ILEC tariff pricing plans of AT&T, CenturyLink, Frontier, and Verizon for special access services.199 In April 2015, the Commission adopted rules for commercial use of 150 megahertz in the 3550-3700 MHz band (3.5 GHz Band), enabling innovative use cases such as wireless backhaul.200 V. PRICING LEVELS AND TRENDS 73. Most U.S. mobile telephone subscribers are billed monthly after their service has been provided (postpaid service). Other U.S. mobile telephone subscribers, including those lacking the necessary credit history, are required to pay for their service in advance (prepaid service). Until 2013, most postpaid subscribers signed a two-year service contract in return for receiving a significant upfront discount on the price of a handset, with service providers recovering the balance of the handset cost over the course of the contract through the higher monthly fees charged for mobile service. Since then, many postpaid subscribers have taken advantage of the option to purchase a handset separately from the service plan via an installment payment plan (equipment 193 See Wireless Infrastructure Report and Order, 29 FCC Rcd at 12867 ¶ 4. 194 See Wireless Telecommunications Bureau Seeks Comment on Revising the Historic Preservation Review Process for Small Facility Deployments, WT Docket No. 15-180, Public Notice, 30 FCC Rcd 8160 (WTB 2015). 195 See Seventeenth Report, 29 FCC Rcd at 15372 ¶ 119. 196 See Special Access for Price Cap Local Exchange Carriers; AT&T Corporation Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, WC Docket No. 05-25, RM- 10593, Order on Reconsideration, 29 FCC Rcd 10899 (WCB 2014) (“Order on Reconsideration”). On August 22, 2012, the Commission had adopted a Report and Order that suspended, on an interim basis, rules that allowed for automatic grants of pricing flexibility for special access services in light of evidence in the record the rules failed to accurately reflect the state of competition in the market for special access. See Special Access for Price Cap Local Exchange Carriers; AT&T Corporation Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, WC Docket No. 05-25, RM-10593, Report and Order, 27 FCC Rcd 10557 (2012). 197 See Special Access for Price Cap Local Exchange Carriers; AT&T Corporation Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, WC Docket No. 05-25, RM- 10593, Order, 29 FCC Rcd 14346 (WCB 2014) (“Extension Order”). 198 See Special Access for Price Cap Local Exchange Carriers; AT&T Corporation Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, WC Docket No. 05-25, RM- 10593, Order, DA 15-1239 (WCB rel. Nov. 2, 2015). 199 See Investigation of Certain Price Cap Local Exchange Carrier Business Data Services Tariff Pricing Plans, WC Docket No. 15-247, Order Initiating Investigation and Designating Issues for Investigation, DA 15-1194 (WCB rel. Oct. 16, 2015). 200 See Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Report and Order and Second Further Notice of Proposed Rulemaking, 30 FCC Rcd 3959, 4024 ¶ 207 (2015). Federal Communications Commission DA 15-1487 53 installment plan, or EIP), leading to a rapid shift from traditional postpaid contract plans to no-contract plans.201 As a result, the requirement to sign a service contract is no longer a key distinction between postpaid and prepaid service. The following discussion of developments in mobile service pricing is divided into two sections – postpaid and prepaid – and focuses on pricing changes during the period covered by this Report. A. Postpaid Service 74. As also discussed in the Seventeenth Report, we have continued to see a range of postpaid pricing changes and promotions.202 In August 2014, Sprint launched shared data plans that offered higher monthly data allowances than similarly-priced shared data tiers then being offered by Verizon Wireless and AT&T, and cut the price of its unlimited data plan for EIP and other non-subsidized subscribers. Sprint’s move to what analysts characterize as a more aggressive pricing strategy, followed a similarly disruptive shift in pricing strategy by T- Mobile beginning in March 2013.203 The following discussion highlights some of the pricing changes and promotions that have been introduced in postpaid service during the period covered by this Report. i. Promotional Pricing Offers and Potential Stabilization 75. Service providers continued to compete for customers by increasing the monthly data allowances on certain tiers of shared data plans while leaving unchanged the existing monthly data charges. Some of these pricing changes were limited-time promotional offers. For example, there was a wave of “double data” promotions in the fall of 2014.204 Then, in August 2015, AT&T launched a new limited-time promotional offer of “15GB for the price of 10GB.”205 Otherwise, service providers used this type of pricing change to restructure their rate plans. In November 2014, AT&T increased the monthly data allowances on two mid-range shared data tiers (from 2GB to 3GB for $40, and from 4GB to 6GB for $70).206 In February 2015, Verizon Wireless added 201 While such offerings have no contracts for the service, cancellation of services before the end of the installment plan period results in a balance due for the unpaid handset charges. 202 See David W. Barden et al., “3Q14 Preview and Model Book – Wireless Pricing is Top of Mind,” Bank of America Merrill Lynch, Equity Research, at 3 (Oct. 17, 2014) (arguing that recent pricing changes reflect “the intersection of T- Mobile and Sprint’s initiatives to gain or sustain subscriber momentum after years of losses, and AT&T and Verizon Wireless’s efforts to hold share and keep churn low.”) 203 See Philip Cusick, “Wireless Competition Intensifies as Verizon, AT&T and Sprint Issue Promotional Price Cuts,” J.P. Morgan, North America Equity Research, at 1 (Nov. 18, 2014); Simon Flannery, et al., “3Q14 Trend Tracker: ‘Tis the Season for Competition and Uncertainty,” Morgan Stanley, Equity Research, at 14 (Nov. 18, 2014). 204 AT&T doubled the data allowances on high-end shared data tiers (15GB to 50GB) through the end of October 2014, and both Verizon Wireless and Sprint followed suit by doubling or commensurately increasing the data allowances on comparable high-end shared data tiers. See David W. Barden, et al., “3Q14 Preview and Model Book – Wireless Pricing is Top of Mind,” Bank of America Merrill Lynch, Equity Research, at 3 (Oct. 17, 2014); “Sprint Doubles the Data of AT&T Promotion with Sprint Family Share Pack and Sprint Business Share Offers,” available at http://newsroom.sprint.com/news- releases/sprint-stands-behind-pledge-to-deliver-double-the-data.htm. AT&T subsequently extended its double data promotion on high-end tiers through mid-November 2014. See “AT&T Adds More Data to New Mobile Share Value Plans on the Nations Most Reliable 4G LTE Network,” available at http://about.att.com/story/att_adds_more_data_to_new_mobile_Share_value_plans_on_the_nations_most_reliable_4g_lte_ne twork.html; John C. Hodulik, et al., “AT&T and Verizon Tweak Pricing, Yet Again,” UBS, Global Research, at 1 (Nov. 3, 2014). 205 See “AT&T Introduces New AT&T Mobile Share Value Plans,” available at http://www.fiercewireless.com/press- releases/att-introduces-new-att-mobile-share-value-plans. See also Phil Goldstein, “AT&T Shakes Up Mobile Share Value Plan Pricing, Increases $100/10 GB plan to 15 GB,” FierceWireless (Aug. 14, 2015). 206 See “AT&T Adds More Data to New Mobile Share Value Plans on the Nation’s Most Reliable 4G LTE Network,” available at http://about.att.com/story/att_adds_more_data_to_new_mobile_Share_value_plans_on_the_nations_most_reliable_4g_lte_ne Federal Communications Commission DA 15-1487 54 1GB of data to the data allowances of its low-end (1GB to 3GB) shared data tiers for the same price as before, in addition to cutting the price of selected higher data tiers and introducing several new data tiers.207 76. Price rivalry in the fall of 2014 also took the form of limited-time promotional pricing offers on selected large data plans, focusing on the four-line (“quad-play”) pricing segment targeted at families.208 The pricing models of the four nationwide service providers are differentiated in this segment. Verizon Wireless, AT&T, and Sprint offer shared data plans in which a single bucket of data can be shared among multiple (up to ten) persons or devices. In contrast, T-Mobile competes in the multiple-line family segment through its family plan offering. T-Mobile’s version of a family plan comes with a dedicated LTE data bucket for each person or device to use on an exclusive basis, but it gives subscribers the option of adding additional lines at reduced prices. After Sprint launched its version of a shared data plan, T-Mobile expanded its family plan offering by increasing the number of lines customers can add from five to ten lines per account, with the price of adding each new line reduced to $10 per month after the second line.209 77. In early November 2014, Verizon Wireless introduced promotional pricing offers of 10GB for $80 per month and 15GB for $100 per month.210 Although Verizon Wireless characterized the offers as a limited- time promotion, no specific end date was set at the time the offer was announced.211 Sprint countered by cutting the prices of its 12GB and 16GB shared data plans to $80 and $100 per month, respectively, specifying that the promotion would be available from November 14, 2014 to January 15, 2015.212 AT&T matched Verizon Wireless’s 15GB for $100 per month offer on November 18, but discontinued this promotion at the end of November.213 In April 2015, Verizon Wireless reintroduced the same two promotional pricing offers, once again with no specified end date.214 78. In late July 2015, T-Mobile ended its existing “10GB for $100” family plan promotion and launched a new limited-time promotional offer that gives each customer 10GB of LTE data per month, with the first two lines priced at $100 per month and each additional line thereafter $20 per month.215 In conjunction with twork.html. See also John C. Hodulik, et al., “AT&T and Verizon Tweak Pricing, Yet Again,” UBS, Global Research, at 1 (Nov. 3, 2014). 207 See “More Options and Even More Value Coming More Everything, available at http://www.fiercewireless.com/press- releases/more-options-and-even-more-value-coming-more-everything; Philip Cusick, “Wireless Pricing Update,” J.P. Morgan, North America Equity Research, at 1 (Feb. 12, 2015). 208 See Philip Cusick, “Wireless Competition Intensifies as Verizon, AT&T and Sprint Issue Promotional Price Cuts,” J.P. Morgan, North America Equity Research, at 1 (Nov. 18, 2014). 209 See “T-Mobile Expands Industry-Leading Family Plan,” available at http://www.fiercewireless.com/press-releases/t- mobile-expands-industry-leading-family-plan-0; Phil Goldstein, “T-Mobile Cuts Prices on Family Plans With 7-10 Lines, Launches $10/Month Tablet Plan,” FierceWireless (Aug. 28, 2014). 210 See John C. Hodulik, et al., UBS, Global Research, at 3 (Nov. 3, 2014). 211 See Ben Rooney, “Verizon is Upping its Game – Offering New Data Plans that Give Smartphone Users More Bang for Their Buck,” CNN Money (Nov. 3, 2014). 212 See Mike Dano, “Sprint Cuts $10 From 12 GB and 16 GB Shared Data Plans,” FierceWireless (Nov. 12, 2014). 213 See Phil Goldstein, “AT&T Cuts Price of 15 GB Shared Data Plan by $30,” FierceWireless (Nov. 18, 2014); Mike Dano, “AT&T Ends 15 GB Promotion After 13 Days,” FierceWireless (Dec. 4, 2014). 214 See Phil Goldstein, “Verizon Reintroduces $80/10 GB and $100/15 GB Shared Data Plans,” FierceWireless (Apr. 24, 2015). 215 See “T-Mobile Amps Up its Family Plan: Family Members get 10GB Each – for Just $30 a Line,” available at http://newsroom.t-mobile.com/news/t-mobile-amps-up-its-family-plan-family-members-get-10gb-each--for-just-30-a- line.htm; Phil Goldstein, “T-Mobile Launches New Family Plan With 10 GB of Data, But Drops Existing Unlimited Promotion,” FierceWireless (July 14, 2015). Federal Communications Commission DA 15-1487 55 an additional promotion allowing customers to add a fourth line free of charge through September 7, 2015, the new plan offered a family of four 40GB of LTE data for $120 per month – more data than the expired promotion but at a higher price point. Sprint quickly countered with two limited-time promotional pricing offers of 10GB of shared data for $100 per month and 40GB of shared data for $120 per month.216 In August 2015, AT&T introduced a revamped version of its shared data pricing plan that included the “15GB for the price of 10GB” promotional offer discussed above.217 Among other pricing changes, AT&T’s new plan also brought its pricing down closer to that of T-Mobile and Sprint for lower data tiers targeted at single-line subscribers.218 Although the “15GB for the price of 10GB” offer was the most heavily promoted pricing change, AT&T characterized the entire package as a “promotional plan” and a “limited-time special offer.”219 In November 2015, T-Mobile announced that customers on their Simple Choice plans would be able to access certain video streaming services without depleting their data allowance.220 Binge-On is similar to T-Mobile’s earlier Music-Freedom promotion,221 and both cover a range of popular free and subscription based streaming services. 79. In addition to its other promotional pricing offers, in December 2014 Sprint launched a promotional offer to cut the monthly bills in half for AT&T and Verizon Wireless customers if they switch to Sprint.222 Though originally scheduled to end in January 2015, the offer was subsequently extended through the end of 2015 and to cover T-Mobile customers.223 As analysts have pointed out, because many customers were still on traditional subsidy plans but the promotion only applied to the service fee component of the customer’s bill, this promotion did not necessarily cut targeted customers’ bills in half when the cost of the handset is taken into account.224 According to one analyst, Sprint conceded that the average customer who signed up via the “Cut Your Bill in Half” promotion actually saw about a 20 percent reduction in their aggregate bill (monthly service fees plus equipment payments) after adjusting for a full device installment or leasing payment.225 Analysts also noted that Sprint’s other promotional pricing offers or shared data plans often offered customers a better deal than the “Cut Your Bill in Half” promotion, so that many of the customers that the promotion succeeded in drawing 216 See “Sprint Introduces Best Plan for Families,” available at http://www.fiercewireless.com/press-releases/sprint- introduces-best-plan-families; Phil Goldstein, “Sprint Targets T-Mobile With New 10 GB/$100, 40 GB/$120 Shared Family Plan,” FierceWireless (July 30, 2015). 217 See “AT&T Introduces New AT&T Mobile Share Value Plans,” available at http://www.fiercewireless.com/press- releases/att-introduces-new-att-mobile-share-value-plans; Phil Goldstein, “AT&T Shakes Up Mobile Share Value Plan Pricing, Increases $100/10 GB Plan to 15 GB,” available at http://www.fiercewireless.com/story/att-shakes-mobile-share- value-plan-pricing-increases-10010-gb-plan-15-gb/2015-08-14, FierceWireless (Aug. 14, 2015). 218 See Philip Cusick, et al., “Verizon Simplifies Pricing and Drops Subsidy Option; AT&T Also Reshuffles Its Pricing,” J.P. Morgan, North America Equity Research, at 1, 3 (Aug. 18, 2015). 219 See “AT&T Introduces New AT&T Mobile Share Value Plans,” available at http://www.fiercewireless.com/press- releases/att-introduces-new-att-mobile-share-value-plans; “AT&T Mobile Share Value Plans,” available at http://www.att.com/shop/wireless/data-plans.html. 220 See “T-Mobile Unleashes Mobile Video with Binge On,” available at https://newsroom.t-mobile.com/media-kits/un- carrier-x.htm. 221 See Seventeenth Report, 29 FCC Rcd at 15386 ¶ 152. 222 See “Sprint’s Offer: Cut Your Wireless Bill in Half Event,” available at http://www.fiercewireless.com/press- releases/sprints-offer-cut-your-wireless-bill-half-event; Phil Goldstein, “Sprint Offers to Cut Monthly Bills in Half for Verizon, AT&T Customers Who Switch,” FierceWireless (Dec. 2, 2014). 223 See William Ho, “Sprint, T-Mobile Will Keep Sparking Pricing Moves in 2015,” FierceWireless (Feb. 25, 2015). 224 See Craig Moffett, “U.S. Wireless: The Race to the Bottom?,” MoffettNathanson Research (June 22, 2015); William Ho, Sprint, T-Mobile Will Keep Sparking Pricing Moves in 2015,” FierceWireless (Feb. 25, 2015). 225 See Craig Moffett, “U.S. Wireless: The Race to the Bottom?,” MoffettNathanson Research (June 22, 2015). Federal Communications Commission DA 15-1487 56 into Sprint retail stores ended up choosing other Sprint plan options instead.226 Sprint has subsequently relabeled the promotion to “Cut Your Rate Plan in Half.”227 80. Despite the wave of promotional pricing activity, analysts have recently noted signs of a possible stabilization of pricing.228 For example, AT&T did not respond to Verizon Wireless’s reintroduction of its promotional pricing offers in April 2015, and although it matched one of the same offers when they were first launched in November 2014, AT&T discontinued the promotion long before Verizon Wireless did.229 In addition, although both Sprint and T-Mobile continue to offer unlimited data plans, one analyst noted that both companies “are deemphasizing unlimited data plans and instead they are promoting their shared data bucket plans,” a trend the same analyst characterized as being “good for profitability.”230 Finally, in August 2015, Verizon Wireless ended its existing promotional pricing offers and launched a greatly simplified version of its shared data plan, called “The Verizon Plan.”231 The effect of Verizon Wireless’s new shared data plan on customers’ monthly bills depends on the size of their data allowance and the number of smartphone lines used.232 Customers on data tiers with less than 6GB of data will see a price cut as a result of reductions in monthly data and smartphone access charges. The price will remained unchanged for customers on the 6GB data tier with two smartphones. Customers on data tiers with a monthly data allowance greater than 6GB and with more than two lines may end up paying slightly higher prices than they had been previously.233 The day after “The Verizon Plan” went into effect, AT&T launched its revamped shared data promotional plans, lowering AT&T’s pricing for low-end single- line data plans and replacing the 10GB plan with its “15GB for the price of 10GB” offer. 81. As summarized by one analyst, following the launches of Verizon Wireless’s and AT&T’s new plans, “Sprint and T-Mobile continue to have the most competitively priced plans at all levels and Verizon 226 See William Ho, “Sprint, T-Mobile Will Keep Sparking Pricing Moves in 2015,” FierceWireless (Feb. 25, 2015); Craig Moffett, “U.S. Wireless: The Race to the Bottom?,” MoffettNathanson Research (June 22, 2015); Joseph Mastrogiovanni, “Sprint Making a Push for Subs into Year-End,” Credit Suisse, Equity Research (Dec. 2, 2014). 227 See “Switch to Sprint and save 50% on Verizon, AT&T or T-Mobile rates,” available at https://www.sprint.com/apps/save50percent/index.html#!/.. 228 See Craig Moffett, et al., “U.S. Wireless: Hell Freezes Over . . . Upgrading Sector to Overweight and Verizon to Buy, Price Target $54,” MoffettNathanson Research, at 11-12 (Aug. 11, 2015); Phil Goldstein, “Analyst: U.S. Wireless Service Revenue Dropped in Q2 But May Be Stabilizing,” FierceWireless (Aug. 12, 2015). 229 See Phil Goldstein, “Analysts: Verizon’s New Plans Could Hurt Net Adds Short-Term, Will Boost Profits Long-Term,” FierceWireless (Aug. 10, 2015). 230 See Phil Goldstein, “Analyst: U.S. Wireless Service Revenue Dropped in Q2 But May Be Stabilizing,” FierceWireless, (Aug. 12, 2015) (quoting Chetan Sharma Consulting). 231 The new plan comes in four basic sizes, ranging from small (1GB for $30 per month) to extra-large (12GB for $80 per month). Larger data buckets are available for customers who need them. In addition to reducing the number of data tiers, Verizon established a uniform monthly access line charge for each smartphone of $20 across the board for EIP subscribers, replacing the previous two-tiered pricing structure in which EIP subscribers paid $25 per smartphone for data tiers with less than 6GB and $15 per smartphone for data tiers with 6GB or more. See “One Plan. Pick a Size. Simple,” available at http://www.verizonwireless.com/landingpages/verizon-plan/?intcmp=INT-MAR-NON-AW-The_Verizon_Plan-081315- 1MPB1-DE-HP-PP-NP; “S-M-L-XL: Choose the Right Size on the Nation’s Best 4G LTE Network,” available at http://www.fiercewireless.com/press-releases/s-m-l-xl-choose-right-size-nations-best-4g-lte-network; Phil Goldstein, “Verizon Cuts Prices on Shared Data Plans and Smartphone Access Charges,” FierceWireless (Aug. 7, 2015). 232 See Phil Goldstein, “Analysts: Verizon’s New Plans Could Hurt Net Adds Short-Term, Will Boost Profits Long-Term,” FierceWireless (Aug. 10, 2015). 233 See id.; Philip Cusick, et al., “Verizon Simplifies Pricing and Drops Subsidy Option; AT&T Also Reshuffles Its Pricing,” J.P. Morgan, North America Equity Research, at 1 (Aug. 18, 2015). Federal Communications Commission DA 15-1487 57 Wireless remains the most expensive option.”234 Finally, effective November 15, 2015, T-Mobile raised prices on its unlimited data plan and two out of three of its usage-based data pricing tiers in conjunction with an increase in the latter two tiers’ monthly high-speed 4G LTE data caps and the launch of Binge On, a free video streaming option that is only available on T-Mobile’s new $65/6GB tier (previously $60/3GB) and above.235 T-Mobile left the price of its base data tier unchanged at $50, but increased the monthly high-speed data cap on this tier as well (from 1GB to 2GB). ii. Device Financing 82. iPhone Promotions. A number of promotional offers during this period were timed to take advantage of Apple’s annual fall release of new iPhone models. In the fall of 2014, rival promotional offers around the iPhone 6/6 Plus release focused on the trade-in value of used devices,236 and also featured the launch of the first handset leasing program. For example, shortly before the fall 2014 iPhone 6/6 Plus release, T-Mobile announced that it would guarantee the best trade-in value on customers’ used devices when they upgrade to a new device,237 and in response, Sprint offered to match any major national competitor’s trade-in value,238 and Verizon Wireless made a similar offer.239 Promotional activity in advance of the fall 2015 iPhone 6s/6s Plus release focused on handset pricing, with handset leasing promotions playing a leading role.240 83. Handset Leasing. The first handset leasing program was Sprint’s “iPhone for Life,” under which monthly payments to lease an iPhone are lower than the monthly installment payments would be under Sprint’s “Easy Pay” EIP, but customers do not get to keep the phone free of charge after two years.241 In June 2015, T- Mobile launched its own version of a handset leasing program called “JUMP! On Demand.”242 T-Mobile subsequently used its new leasing program to launch new promotions featuring discounted monthly leasing 234 Philip Cusick, et al., “Verizon Simplifies Pricing and Drops Subsidy Option; AT&T Also Reshuffles Its Pricing,” J.P. Morgan, North America Equity Research, at 1 (Aug. 18, 2015). 235 Philip Cusick, et al., “Wireless Pricing Update: T-Mobile Raises Prices While Sprint Now Offers Unlimited 2G Data; Binge On seems Fine,” J.P. Morgan, North America Equity Research (Nov. 16, 2015). 236 See Philip Cusick, “Carriers Now Offer Competitive Trade-in Values Compared to Retailers for Used Smartphones,” J.P. Morgan, North America Equity Research, at 1 (Oct. 27, 2014). 237 See “T-Mobile Guarantees Industry’s Best Trade-In Value on Used Devices,” available at http://newsroom.t- mobile.com/news/t-mobile-guarantees-industrys-best-trade-in.htm; Mike Dano, “T-Mobile Launches Phone Trade-in Guarantee, Promises to Beat Any Offer by $50,” FierceWireless (Sept. 8, 2014). 238 See Phil Goldstein, “Sprint Hits Back at T-Mobile With its Own Phone Trade-in Guarantee,” FierceWireless (Sept. 9, 2014). 239 See Philip Cusick, “Carriers Now Offer Competitive Trade-in Values Compared to Retailers for Used Smartphones, J.P. Morgan, North America Equity Research, at 1 (Oct. 27, 2014). 240 See John C. Hodulik, et al., “Kicking Up Competition for Handset Pricing,” UBS, Equity Research (Sept. 24, 2015). 241 See “iPhone 6s,” available at http://www.sprint.com/landings/iphone/index.html#whysprint?ECID=vanity:iphone; Philip Cusick, “3Q14 Preview and Industry Update,” J.P. Morgan, North America Equity Research, at 9 (Oct. 20, 2014). Sprint later expanded the handset leasing program to include other high-end 4G smartphones. See also Philip Cusick, “Wireless Pricing Update,” J.P. Morgan, North America Equity Research, at 1 (Feb. 12, 2015). 242 See “T-Mobile Unveils All-New “Jump! On Demand” – a Whole New Way to Get a Phone Whenever You Want,” available at http://newsroom.t-mobile.com/news/t-mobile-unveils-all-new-jump-on-demand--a-whole-new-way-to-get-a- phone-whenever-you-want.htm; Simon Flannery, “UnCarrier Amped Reinforces Competitive Environment,” Morgan Stanley, Equity Research (June 26, 2015); Jennifer M. Fritzsche, “TMUS: Launches First Un-Carrier Amped Initiative – JUMP! On Demand,” Wells Fargo, Equity Research (June 25, 2015). Federal Communications Commission DA 15-1487 58 payments for the next iPhone model, first in July 2015 and then again in September 2015.243 Sprint quickly countered both promotions.244 In the second round, T-Mobile offered to lease a new iPhone 6s for $5 per month ($9 per month for the iPhone 6s Plus) with trade-in of an iPhone 6 or other qualifying smartphone, but with eligibility limited to customers who already own their current device. Sprint countered with a monthly leasing rate of $1 per month ($5 per month for the 6s Plus) under similar terms.245 84. While the low monthly leasing payments have the appearance of a price cut, the discounts to monthly leasing payments effectively replace the provider’s standard upfront payment for the trade-in value of the customer’s existing device, and in T-Mobile’s case the cumulative monthly discount ($15) over the 18-month leasing term amounts to less than the $350 upfront payment the provider previously offered for an iPhone 6 trade- in.246 Absent a handset leasing plan, Verizon Wireless responded with a new program that allows customers to upgrade to a new iPhone every year with trade-in of their current device, provided they have paid off more than half the cost of the current device and purchase the new phone through Verizon Wireless’s EIP.247 Under Verizon Wireless’s non-promotional upgrade policy, EIP customers are required to pay off the full cost of their old device before upgrading, though they are able to upgrade at any time without a trade-in. 85. Tablet Subsidies and Promotions. The growth of tablet subsidies to encourage subscribers to use tablets on mobile networks as well as Wi-Fi continued during the period covered by this Report.248 At various times, service providers offered significant discounts on selected tablets, and in some cases, service providers practically gave certain tablet models away free of charge, to qualifying customers (e.g., those who sign a two- year service contract, sign up for a qualifying data plan, or purchase a smartphone through an EIP program.)249 Service providers also offered discounts on other types of devices, including wearables and accessories. This type of promotional activity is particularly intense during the holiday shopping season. iii. Efforts to Promote No-Contract Service Plans 86. The Seventeenth Report discussed the rise of the EIP, together with the no-contract service plan, as an alternative to the traditional handset subsidy model.250 Beginning with T-Mobile, all four nationwide service providers and some regional providers introduced versions of an EIP and encouraged customers to adopt EIPs through two types of incentives. First, providers offered early handset upgrade options that allow customers to upgrade earlier and more frequently than under traditional subsidized plans but require customers who take advantage of this early upgrade option to pay for their new handsets with an EIP. Second, providers offered lower 243 See Phil Goldstein, “T-Mobile Offers iPhone 6s for as Low as $5/Month in Device Payments if Customers Trade in an iPhone 6,” FierceWireless (Sept. 23, 2015); Phil Goldstein, “T-Mobile Lets Customers Get Free Upgrade to Next iPhone, Lock in $15 Monthly Fee,” FierceWireless (July 28, 2015). 244 See Phil Goldstein, “Sprint Undercuts T-Mobile, Offers iPhone 6s for $1/Month in Leasing Payments With an iPhone Trade-In,” FierceWireless (Sept. 24, 2015); “Sprint Customers Can Upgrade Their iPhone Anytimes, Included in Their Monthly Rate,” available at http://newsroom.sprint.com/news-releases/sprint-customers-can-upgrade-their-iphone-anytime- included-in-their-monthly-rate.htm; Phil Goldstein, “Sprint Hits Back at T-Mobile With Its Own $15/Month iPhone Promotion, Says Customers Can Upgrade to a New iPhone Any Time,” FierceWireless (Aug. 17, 2015). 245 See John C. Hodulik, et al., “Kicking Up Competition for Handset Pricing,” UBS, Equity Research (Sept. 24, 2015). 246 See id. 247 See Phil Goldstein, “Apple’s iPhone 6s, 6s Plus Go On Sale as Verizon Looks to Match Sprint, T-Mobile Offers,” FierceWireless (Sept. 25, 2015). 248 See Phil Goldstein, “As U.S. Tablet Sales Fall Overall, Verizon, AT&T and Sprint See Increasing Sales of Cellular Tablets,” FierceWireless (May 14, 2015). 249 See Phil Goldstein, “Verizon, AT&T, Sprint, T-Mobile and More Unleash Device Deals for Black Friday,” FierceWireless (Nov. 26, 2014). 250 See Seventeenth Report, 29 FCC Rcd at 15381 ¶ 140. Federal Communications Commission DA 15-1487 59 monthly service fees to customers who sign a contract to pay for a new handset in monthly installments, as well as customers who pay the full price for their handset upfront, new customers who bring their own devices (“BYOD”), and existing customers who have already paid for their handset under a traditional contract plan that has expired. 87. In the period since the Seventeenth Report, providers continued to adjust the incentives offered to encourage subscribers to choose EIPs and no-contract service plans over the traditional handset subsidy option. For example, in a series of pricing changes from October 2014 to August 2015, Verizon Wireless further reduced the already discounted monthly access line charges that EIP subscribers on small and medium-sized data plans pay for each smartphone.251 Sprint waived the monthly access line charges for new subscribers on qualifying data plans in conjunction with some of its promotions, but only if subscribers purchased their devices through its EIP or handset leasing options or brought their own.252 The introduction of handset leasing programs can also be viewed as a way of encouraging subscribers to shift to no-contract plans. To the extent subscribers find the terms of these handset leasing programs more attractive than the terms of existing EIP programs, the launch of handset leasing programs by Sprint and T-Mobile may enhance incentives to choose no-contract service plans. 88. Service providers also moved beyond the use of incentives to promote the shift to no-contract plans. During the period covered by the Seventeenth Report, T-Mobile remained the only nationwide provider that had ceased offering handset subsidies. Although the other nationwide service providers promoted EIPs and no-contract plans to varying degrees, all three continued to offer customers the traditional contract plan with handset subsidy option. Since then, the remaining nationwide providers have taken steps to limit sales of contract plans and bring an end to handset subsidies or have announced plans to do so. AT&T opted to limit sales of subsidized handsets through third-party retail partners such as Walmart and Apple. Beginning in June 2015, third-party national retailers and local authorized dealers progressively discontinued selling two-year service contracts with subsidized handsets to AT&T customers, and instead required them to purchase handsets at full retail price or through AT&T’s EIP.253 For the time being, however, AT&T will continue to offer two-year service contracts with subsidized handsets in its company-owned retail stores and through its online sales channels. Significantly, the revamped Mobile Share Value promotional plans that AT&T introduced in August 2015 included the option to purchase a subsidized smartphone through a two-year service contract.254 89. Verizon Wireless moved in the same direction with the introduction of its new shared data plan “The Verizon Plan” in August 2015. Unlike Verizon Wireless’s previous shared data plan, the new plan does not include the option to purchase a handset at a discounted price through a two-year service contract. All new customers and also existing customers who move to the new plan will buy or upgrade to new smartphones either by using Verizon Wireless’s EIP device payment option or by paying the full retail price upfront.255 Note that like 251 See John C. Hodulik, et al., “AT&T and Verizon Tweak Pricing, Yet Again,” UBS, Global Research, at 3 (Nov. 3, 2014); “More Options and Even More Value Coming for More Everything,” available at http://www.fiercewireless.com/press- releases/more-options-and-even-more-value-coming-more-everything; “S-M-L-XL: Choose the Right Size on the Nation’s Best 4G LTE Network,” available at http://www.fiercewireless.com/press-releases/s-m-l-xl-choose-right-size-nations-best- 4g-lte-network. 252 See “Sprint Stands Behind Pledge to Deliver ‘Double the Data,’” available at http://newsroom.sprint.com/news- releases/sprint-stands-behind-pledge-to-deliver-double-the-data.htm; “Sprint Introduces Best Plan for Families,” available at http://www.fiercewireless.com/press-releases/sprint-introduces-best-plan-families. 253 See Phil Goldstein, “Report: AT&T to Abandon 2-Year Contracts at National Retailers and Local Dealers,” FierceWireless, May 20, 2015; Phil Goldstein, “Report: Apple Stores to Sell Only Unsubsidized AT&T iPhones, Will Adopt Verizon Edge Changes,” FierceWireless (June 2, 2015). 254 See Philip Cusick, et al., “Verizon Simplifies Pricing and Drops Subsidy Option; AT&T Also Reshuffles Its Pricing,” J.P. Morgan, North America Equity Research at 1 (Aug. 18, 2015). 255 See “S-M-L-XL: Choose the Right Size on the Nation’s Best 4G LTE Network,” available at http://www.fiercewireless.com/press-releases/s-m-l-xl-choose-right-size-nations-best-4g-lte-network. Federal Communications Commission DA 15-1487 60 AT&T, existing Verizon Wireless subscribers on two-year contracts will not be required to switch to “The Verizon Plan” even to upgrade their phone, but instead will be able to keep and renew their existing contract plan and continue to take advantage of subsidized devices offered by Verizon Wireless when they upgrade to a new smartphone.256 Finally, in August 2015, Sprint CEO Marcelo Claure disclosed that Sprint will no longer offer two-year contracts and handset subsidies by the end of 2015 and will move entirely to the device leasing model.257 iv. Early Termination Fee Buyouts and Other Switching Incentives 90. With the shift away from handset subsidies, marketing tactics have increasingly focused on Early Termination Fee (“ETF”) buyouts to encourage customers to switch from rivals. ETF buyouts typically include a cash payment or credit to reimburse ETFs for customers on traditional contract plans, or alternatively, to pay off the remaining balance of an EIP, plus a separate device credit for trading in a customer’s current handset.258 For example, T-Mobile maintained its offer to reimburse up to $650 in ETFs for customers switching from the other three nationwide service providers since it was first introduced in January 2014.259 In May 2015, T-Mobile supplemented this offer with its “Never Settle” free two-week trial promotion targeted specifically at Verizon Wireless customers.260 Under the promotion, Verizon Wireless customers port their number to a new T-Mobile smartphone while holding on to their Verizon Wireless phone. If customers decide to switch permanently to T- Mobile after the trial, T-Mobile will reimburse them for up to $650 in Verizon’s ETFs and outstanding device payments, provided customers trade in their Verizon Wireless phone and purchase a new phone from T-Mobile.261 91. Rival service providers tended to rely on a series of limited-time ETF buyout offers, often timed to coincide with the launch of other types of pricing changes and promotions. For example, Verizon Wireless offered a one-time $100 bill credit per qualified line to new EIP customers who port in their wireless number from another service provider when it restructured the pricing of its shared data offerings in February 2015.262 Sprint’s “Cut Your Bill in Half” offer for customers switching from Verizon Wireless and AT&T, as described above, included an ETF buyout of up to $350 per line to reimburse any remaining balance on their EIP as well as early termination fees.263 From January 23 through April 9, 2015, Sprint guaranteed customers who switch from T- Mobile a minimum trade-in value of $200 for their used smartphones in addition to up to $350 per line to cover switching costs.264 One analyst notes that, while ETF switching credits are “an essential component to subscriber 256 See “Simplified Data Choices Match Customer Lifestyles,” available at http://www.verizonwireless.com/news/article/2015/08/simplified-data-choices-match-customer-lifestyles.html; Phil Goldstein, “Verizon: Existing Customers With 2-Year Contracts and Subsidized Smartphones Can Keep Them,” FierceWireless (Aug. 31, 2015). 257 See Phil Goldstein, “Sprint to Abandon 2-Year Contracts By Year-End, Embrace Leasing Exclusively,” FierceWireless (Aug. 17, 2015). 258 See Seventeenth Report, 29 FCC Rcd at 15382 ¶ 145. 259 See “Switch Carriers: No Early Termination Fee,” available at http://www.t-mobile.com/offer/switch-carriers-no-early- termination-fee.html?icid=WMD_TM_Q115CRRRFR_LNN5NV72F4X1691. 260 See Phil Goldstein, “T-Mobile Targets Verizon Customers With Free 2-Week Trial,” FierceWireless (May 5, 2015). 261 See Phil Goldstein, “T-Mobile Extends ‘Never Settle’ Free 2-Week Trial With Promotion Targeting Verizon,” FierceWireless (June 3, 2015). 262 See “More Options and Even More Value Coming for More Everything,” available at http://www.fiercewireless.com/press-releases/more-options-and-even-more-value-coming-more-everything. 263 See Phil Goldstein, “Sprint Offers to Cut Monthly Bills in Half for Verizon, AT&T Customers Who Switch,” FierceWireless (Dec. 2, 2014). 264 See Phil Goldstein, “Sprint Offers $200 Trade-in Credit, $350 to Cover ETFs for Switching T-Mobile Customers,” FierceWireless (Jan. 23, 2015). Federal Communications Commission DA 15-1487 61 growth” for T-Mobile, Sprint, AT&T, and Verizon Wireless, they “have tended to play around with lesser (e.g., $100-$150) values.”265 v. Rollover Data Programs 92. One novel pricing change during this period was the launch of rollover data programs enabling mobile subscribers on usage-based data plans to roll over unused data from their monthly data allowances from month to month. Although the practice of allowing mobile subscribers to roll over unused voice minutes from their monthly voice buckets from month to month was first introduced by Cingular Wireless (predecessor to AT&T) in 2002, the application of the rollover concept to unused data was new. Regional operator C Spire was the first service provider to apply the rollover concept to unused data. In November 2014, C Spire launched several rollover data plans that let subscribers carry over unused data to the following month, and touted the plans as a means of eliminating surprise data overage charges.266 In January 2015, T-Mobile launched a rival rollover data program called “Data Stash” that is automatically available to new and existing postpaid subscribers on qualifying data plans at no additional charge, and offered to deposit 10GB of data free of charge in every customer’s Data Stash through the end of 2015.267 AT&T launched its own version of the rollover data concept shortly thereafter, limiting its availability to EIP and other non-subsidized subscribers on no-contract shared data plans.268 93. One of the differentiating features of rival rollover data programs is how long unused data remains available for use. Under T-Mobile’s program, subscribers can use data any time for up to one year from the date it is deposited into their Data Stash, but after that date they will lose any of their unused data. Under AT&T’s program, any unused data in a given month automatically rolls over and is available to be used within the next month. Within a given month, customers use their monthly shared data allowance first, before they begin to use any unused rollover data from the previous month, and any unused rollover data from the previous month does not carry over to the following month. B. Prepaid Service 94. The four nationwide service providers offer prepaid service under their own prepaid brands, in addition to selling mobile wireless service wholesale to MVNOs, which then resell service on the nationwide networks under other prepaid brands. Prepaid strategies vary across providers.269 Verizon Wireless, the nationwide service provider with the smallest share of prepaid subscribers, has one prepaid brand with what one analyst characterizes as a “thin portfolio” of offerings.270 In contrast, the largest U.S. reseller, Tracfone, has multiple prepaid brands that target different market and demographic segments. To varying degrees, the other three nationwide service providers pursue a multi-brand prepaid strategy similar to that of Tracfone. 95. One factor that used to differentiate prepaid service from postpaid was the presence of multi- regional prepaid service providers, such as MetroPCS and Leap (which offered service under the Cricket brand). 265 See William Ho, “Sprint, T-Mobile Will Keep Sparking Pricing Moves in 2015,” FierceWireless (Feb. 25, 2015). 266 See “C Spire Debuts Customer-Inspired Rolling Data Plans that Promise to Save Consumers Money and Eliminate Surprise Data Overages,” available at http://www.cspire.com/company_info/about/news_detail.jsp?entryId=22400005. 267 See “”Data Stash: Don’t Lose What You Don’t Use,” available at http://www.t-mobile.com/offer/data-stash-data- roll.html; Phil Goldstein, “T-Mobile Launches Rollover Data Program, Gives Customers 10GB for Free to Start,” FierceWireless (Dec. 16, 2014). Eligibility was limited to subscribers on qualifying postpaid Simple Choice plans who have purchased at least 3GB of 4G LTE data for smartphones, or 1GB or more of 4G LTE data for postpaid tablet plans. 268 See “AT&T Gives More Than 50 Million Mobile Share Value Subscribers Shareable Rollover Data,” available at http://about.att.com/story/att_gives_50_million_customers_shareable_rollover_data_at_no_charge.html; “Rollover Data,” available at http://www.att.com/shop/wireless/rollover-data.html. 269 See William Ho, “Is Prepaid Still Relevant? Yes!,” FierceWireless (Oct. 27, 2014). 270 See id. Federal Communications Commission DA 15-1487 62 During the period covered by the Seventeenth Report, T-Mobile acquired MetroPCS and AT&T acquired Leap. As one analyst put it, the prepaid segment has become “a big-4 game.”271 Analysts sub-divide prepaid service offerings into a high-end segment for the recurring monthly plan user and the traditional segment with pay-as- you-go pricing. Like postpaid service plans, prepaid monthly plans are typically offered in multiple tiers with different rates and usage levels. Prepaid offerings include pay-as-you-go pricing and “plans with features that mimic postpaid.”272 In this regard, one noteworthy trend is an increase in prepaid subscribers switching to postpaid service in response to more attractive postpaid plan pricing and the recent postpaid pricing changes and promotions.273 Analysts believe the option available to postpaid subscribers to purchase a handset on an installment payment plan with no down payment is an added attraction for prepaid subscribers because prepaid subscribers ordinarily do not have access to the device financing options available to postpaid subscribers, but instead have to pay the full price of handsets upfront. 96. The migration of former prepaid subscribers to postpaid service in response to postpaid pricing and promotional changes is important background to changes in prepaid plan pricing in the period since the Seventeenth Report. To a significant extent, service providers implemented the same or similar types of pricing changes and promotions in their high-end prepaid monthly service offerings as they did in their postpaid service offerings. In some cases, pricing changes and promotions originally launched for postpaid service were subsequently extended to one or more prepaid brands, while in other cases, the prepaid brand was the first to launch a particular type of pricing change. The list of pricing and promotional changes common to both postpaid and high-end prepaid service offerings includes additions to data for the same monthly rate (including “double data” promotions), the “cut your bill in half” promotion, switching incentives, shared data plans, rollover data programs, and handset financing options. As explained below, however, the prepaid and postpaid versions of a given pricing change or promotion often differ due to the remaining distinctions between postpaid and prepaid service. The remaining differences largely reflect the different characteristics of postpaid and prepaid subscribers: “prepaid subscribers are typically prepaid for a reason, relating to their income and credit.”274 To prevent credit losses and mitigate the credit risk associated with the prepaid segment, service providers require advance payment for both prepaid service and handsets, and as explained below, most prepaid service providers impose speed reductions for data usage in excess of monthly data allowances, rather than the overage charges that are typical of postpaid data plans. i. More Data for the Same Price and Other Pricing Offers 97. On numerous occasions, prepaid service providers added more data for the same monthly rate to service plans, either on a limited-time promotional basis or in the form of permanent changes to non-promotional rate plans. In the context of prepaid service, this type of pricing change generally consists of an addition to the subscriber’s monthly high-speed data allowance (typically either LTE or HSPA+). As a general rule, prepaid subscribers who reach the limit of their high-speed data allowance in a given month may continue to use their handsets for data service on an unlimited basis, but at reduced speeds.275 In contrast, postpaid subscribers who use 271 See Philip Cusick, “3Q14 Preview and Industry Update,” J.P. Morgan, North America Equity Research, at 13 (Oct. 14, 2014). 272 See William Ho, “Is Prepaid Still Relevant? Yes!,” FierceWireless (Oct. 27, 2014). 273 See id.; Philip Cusick, et al., “Postpaid Outperformed From Tablets While Prepaid Underperformed From Migration to Postpaid and Lifeline Recertification,” J.P. Morgan, North America Equity Research, at 1-2 (May 28, 2015); David W. Barden, et al., “3Q14 Preview and Model Book – Wireless Pricing is Top of Mind,” Bank of America Merrill Lynch, Equity Research, at 7 (Oct. 17, 2014). 274 See David W. Barden, et al., “3Q14 Preview and Model Book – Wireless Pricing is Top of Mind, Bank of America,” Merrill Lynch, Equity Research, at 7 (Oct. 17, 2014). “The prepaid segment churn is as much as 3x higher than postpaid and it is by definition 100% voluntary. The reason for the elevated churn rate here is that prepaid customers may not be able to make consistent payments or they are quick to churn off for the next, lower-priced deal at another carrier.” Id. 275 See Seventeenth Report, 29 FCC Rcd at 15387, 15390 ¶¶ 155, 159. Federal Communications Commission DA 15-1487 63 up their plan’s data allowance in a given month generally incur overage charges if they exceed the allowance.276 T-Mobile, an exception, has eliminated overage charges as part of its “uncarrier” strategy and generally uses speed reductions to control the data usage of both postpaid and prepaid subscribers.277 98. Prior to the wave of “double data” promotions in postpaid service in the fall of 2014, there was a similar wave of promotions in prepaid service. For example, in September 2014, Sprint’s Boost Mobile launched a limited-time promotion in which it doubled the monthly data allowances of its plans and cut the price by $5 per month.278 Shortly thereafter, AT&T’s Cricket launched a rival promotion, offering double the usual monthly data allowance for the same price on two plans and a smaller increase in high-speed data for the same price on a third plan.279 In November 2014, Boost Mobile renewed its promotion, and T-Mobile’s GoSmart prepaid unit likewise doubled the data allowances for several of its rate plans.280 In late December 2014, Boost Mobile announced that its double data promotional pricing would now be its permanent pricing.281 Subsequently, GoSmart and Cricket implemented similar types of pricing changes, in both cases increasing the high-speed data allowances of their existing rate plans.282 Other prepaid service providers also added more data to their rate plans during the same period. In September 2014, T-Mobile’s MetroPCS doubled the data allowance on its entry-level plan and increased the data allowance on its next highest tier.283 The following January, MetroPCS announced that every customer with an LTE data bucket would automatically get an additional gigabyte of LTE data for the same monthly rate.284 Tracfone increased the monthly data allowances of rate plans for its Simple Mobile285 and Straight Talk prepaid brands.286 99. In June 2015, Sprint’s Boost Mobile launched a “Slash Your Payment in Half” promotion targeted at MetroPCS and Cricket customers.287 From June 19 through July 20 of that year, Boost offered to cut those customers’ monthly service bills in half for a full year if they switch to Boost and port their phone number. After the 12-month period expires and the introductory rate ends, customers will automatically be migrated to a 276 See id., 29 FCC Rcd at 15385 ¶¶ 148-49. 277 See id., 29 FCC Rcd at 15385 ¶ 151. 278 See Phil Goldstein, “Sprint’s Boost Mobile Cuts Prices, Doubles Data on Prepaid Plans,” FierceWireless (Sept. 2, 2014); “Boost Mobile Introduces ‘Data Boost Plans’ That Offer Big Data at a Small Price,” available at http://www.fiercewireless.com/press-releases/boost-mobile-introduces-data-boost-plans-offer-big-data-small-price-0. 279 See “Cricket Offers Customers More High-Speed Data in Unlimited Plans with Taxes and Fees Included,” available at http://cricketwireless.mediaroom.com/index.php?s=20295&item=122435. 280 See Phil Goldstein, “Sprint’s Boost, T-Mobile’s GoSmart Offer New Double-Data Promotions,” FierceWireless (Nov. 5, 2014). 281 See Phil Goldstein, “Sprint’s Boost Makes Promotional Plans Permanent, Launches ZTE Speed,” FierceWireless (Dec. 23, 2014). 282 See Phil Goldstein, “T-Mobile’s GoSmart and AT&T’s Cricket Add Data to Plans, U.S. Cellular Drops Prepaid Pricing,” FierceWireless (Jan. 30, 2015). 283 See David W. Barden, et al., 3Q14 Preview and Model Book – Wireless Pricing is Top of Mind, Bank of America Merrill Lynch, Equity Research, at 3 (Oct. 17, 2014). 284 See Phil Goldstein, “T-Mobile’s MetroPCS, Verizon Beef Up Data Offerings for Prepaid Plans,” FierceWireless (Jan. 21, 2015). 285 See Dennis Bournique, “Simple Mobile Adds More Data to Most Plans,” Prepaid Phonenews (Jan. 23, 2015). 286 See “Tracfone’s Straight Talk Boosts Monthly Data Bucket to 5 GB; Nvidia’s Forecast Depressed on Sluggish PC Sales,” FierceWireless (May 8, 2015). 287 See “Boost Mobile Invites All Cricket and MetroPCS Customers to Slash Their Payment in Half,” available at http://newsroom.sprint.com/article_display.cfm?article_id=11916; Phil Goldstein, “Sprint’s Boost Mobile Offers to Cut MetroPCS, Cricket Customers’ Bills in Half If They Switch,” FierceWireless (June 19, 2015). Federal Communications Commission DA 15-1487 64 comparable higher-priced plan. Compared with Sprint’s “Cut Your Bill in Half” promotion for postpaid service, Boost’s promotion resulted in much larger effective percentage discounts on prepaid service because there are no device subsidies and prepaid devices are purchased at full price upfront.288 100. Switching Incentives. Service providers also offered incentives to encourage their rivals’ subscribers to switch to their prepaid brands. The types of switching incentives offered included $100 switching or porting credits and one month of free service.289 For example, AT&T’s Cricket began offering a $100 bill credit for T-Mobile and MetroPCS subscribers who switch to Cricket in August 2014,290 and then renewed the promotion in December 2014, but expanded the offer to include subscribers switching from a wider range of service providers, including Cincinnati Bell Wireless (which later exited the wireless business), Sprint’s Boost Mobile and Virgin Mobile prepaid brands, and also Sprint postpaid service.291 In April 2015, Cricket offered customers who switch from MetroPCS, Boost Mobile, and other service providers one month of service free of charge after completing two months of service.292 Similarly, at various times in late 2014 and the first quarter of 2015, Sprint’s Boost Mobile offered a $100 credit or discount for the purchase of a new device to customers who switch and port their number to Boost from other service providers, or the first month of service free of charge for new customers who do not port their number from another service provider.293 ii. Shared Data Plans and Rollover Data 101. Shared Data Plans. In January 2015, Sprint’s Virgin Mobile launched “Data Done Right,” the first shared data plans for prepaid subscribers.294 Like the postpaid version, the monthly data allowance of Virgin’s shared data plans can be shared by multiple persons, and each tier also includes unlimited talk and text. In other respects, the prepaid version of shared data plans is structured differently than its postpaid counterpart. Under postpaid shared data plans, there is a monthly fee for the data allowance, and a separate monthly line access fee for each smartphone or other device added to the account, with a limit of up to ten devices per account. Under Virgin’s prepaid version, there are no additional monthly line access fees. Instead, there is a set limit on the number of lines for each data tier (two, three, or four depending on the size of the data allowance), and a single monthly fee that covers the cost of both the data allowance and the multiple lines. The plan also features a preloaded app that enables customers to manage data sharing directly from their device. 102. Rollover Data Programs. Following the launch of rollover data programs for postpaid service, several service providers launched prepaid versions of rollover data programs. In March 2015, T-Mobile extended its “Data Stash” rollover data program to its prepaid customers on qualifying data plans, including its offer to start off every subscriber’s Data Stash with a deposit of 10GB of LTE data free of charge through the end 288 See Craig Moffett, “U.S. Wireless: The Race to the Bottom?,” MoffettNathanson Research (June 22, 2015). 289 See William Ho, “Is Prepaid Still Relevant? Yes!,” FierceWireless (Oct. 27, 2014). 290 See Phil Goldstein, “AT&T’s Cricket Targets T-Mobile and MetroPCS Subs With $100 Credit to Switch,” FierceWireless, (Aug. 22, 2014); “Cricket Offers Customers More High-Speed Data in Unlimited Plans with Taxes and Fees Included,” available at http://cricketwireless.mediaroom.com/index.php?s=20295&item=122435. 291 See Phil Goldstein, “AT&T’s Cricket Takes Aim at Sprint, T-Mobile and Cincinnati Bell with $100 Credit,” FierceWireless (Dec. 11, 2014). 292 See Phil Goldstein, “AT&T’s Cricket Launches New Device Financing Options, Kills Promotional 20 GB Plan,” FierceWireless (Apr. 8, 2015). 293 See Phil Goldstein, “AT&T’s Cricket Takes Aim at Sprint, T-Mobile and Cincinnati Bell with $100 Credit,” FierceWireless (Dec. 11, 2014); Phil Goldstein, “Sprint’s Boost Pumps Up Data from 1 GB to 2.5 GB for $35/Month Plan With Auto-Pay,” FierceWireless (Jan. 23, 2015). 294 See “Virgin Mobile USA to Offer Industry’s First No-Contract Data Sharing Plans Only at Walmart,” available at http://newsroom.sprint.com/news-releases/virgin-mobile-usa-to-offer-industrys-first-no-contract-data-sharing-plans-only-at- walmart.htm. Federal Communications Commission DA 15-1487 65 of 2015.295 In May 2015, AT&T started letting GoPhone customers who own a smartphone and have signed up for a qualifying data plan access the service provider’s rollover data program provided they renew their plan on time every month.296 In June 2015, Sprint launched a rollover data program for customers of its Sprint Prepaid brand.297 Under the “Monthly Rolling Data” program, Sprint prepaid customers on certain plans can roll over unused data from month to month and store up to 30GB of data with no expiration date. The offer was made available exclusively through Best Buy for a limited time. iii. Handset Financing 103. Both traditional subsidized contract plans and no-contract EIP plans give postpaid subscribers the option to reduce the upfront cost of purchasing a new handset by spreading the cost of the handset over a specified period of time. In contrast, prepaid subscribers typically have to pay in full for the device upfront. This difference in the availability of device financing tends to limit the options available to prepaid subscribers. In the words of one analyst, “While halo devices such as the iPhone are offered in prepaid, the bread and butter devices sub-$100 draw the most attention. These price points are usually populated with less known brands, such as Alcatel One Touch and ZTE, or lower-end models of global brands not seen in postpaid.”298 In April 2015, AT&T’s Cricket launched three new credit financing options that enable prepaid subscribers to obtain a smartphone for a small upfront payment.299 Two of the options are versions of an installment loan, both requiring a credit check. The third is a rent-to-own leasing option. The leasing option does not require a credit check, though a third party may be used to verify financial information, but does require a larger upfront payment than the installment loan options. C. Price Indicators for Mobile Wireless Services 104. As the discussion above shows, there is wide variety of pricing plans offered by the different mobile wireless service providers that vary along several dimensions, and that may frequently change.300 As discussed earlier301 and in previous Reports, it is difficult to identify sources of information that track actual mobile wireless service prices in a comprehensive and consistent manner. Chart V.C.1 below presents monthly postpaid prices for the four nationwide service providers. Table V.C.1 below shows the current monthly prepaid prices. As reflected in these two charts, sharing data among different lines in an account is a common feature of postpaid pricing plans, but is not an option for prepaid plans. 295 See Phil Goldstein, “T-Mobile Brings ‘Data Stash’ Rollover Data Plan to Prepaid Customers,” FierceWireless (Mar. 16, 2015). 296 See Phil Goldstein, “AT&T Brings Rollover Data to Some of its GoPhone Prepaid Customers,” FierceWireless (May 11, 2015). 297 See Phil Goldstein, “Sprint Unveils Rollover Data for Some Prepaid Customers, Boost Mobile TV Service Geared Toward Sports,” FierceWireless (June 2, 2015). 298 See William Ho, “Is Prepaid Still Relevant? Yes!,” FierceWireless (Oct. 27, 2014). 299 See Phil Goldstein, “AT&T’s Cricket Launches New Device Financing Options, Kills Promotional 20 GB Plan,” FierceWireless (Apr. 8, 2015); “New Cricket Phone Payment Plans Help Customers Get the Phone They Want When They Want It – Including the Upcoming Galaxy S6,” available at http://www.fiercewireless.com/press-releases/new-cricket-phone- payment-plans-help-customers-get-phone-they-want-when-the. 300 See Appendix V. which provides additional information, based on data from RBC Capital Markets, on pricing over time for the four nationwide service providers. 301 See section II.D.1 supra. Federal Communications Commission DA 15-1487 66 Note: The above prices were taken from service providers’ websites on Oct. 21, 2015. Prices include any per line charges indicated by the provider. Prices do not include any additional charges such as equipment installment plans, insurance, international use, or data overage; however, if a service provider includes any such feature in its basic plan, the above price would include this feature. Further, the above prices do not include any one-time charges paid, such as activation fees and termination fees. Prices and the specifics of the plans are subject to change. Table V.C.1 Monthly Prepaid Prices for Major Service Providers Provider GB per line 1 line 2 lines 4 lines Cricket (AT&T) 2.5 $40 $70 $130 Boost (Sprint) 2 $35 $70 $140 Metro (T-Mobile) 2 $40 $75 $145 Straight Talk 5 $45 $90 $180 Note: The above prices were taken from service providers’ websites on Oct. 21, 2015. Prices include any per line charges indicated by the service provider. Prices and the specifics of the plans are subject to change. $0 $20 $40 $60 $80 $100 $120 $140 $160 $180 1 line / 3 GB 2 lines / 6 GB 4 lines / 12 GB Chart V.C.1 Monthly Postpaid Prices for Top 4 Service Providers AT&T Sprint* T-Mobile Verizon Wireless *Sprint's price reflects unlimited for 1 line, 8 GB for 2 lines, and 10 GB for 4 lines. Federal Communications Commission DA 15-1487 67 VI. NON-PRICE RIVALRY 105. Mobile wireless service providers also compete for customers on dimensions other than price, including investment, capacity, network coverage and technology, service quality, and advertising and marketing. Service providers take these actions in an attempt to differentiate themselves from competitors, as well as to adopt certain initiatives of their competitors that have been successful in attracting customers. Such non-price rivalry can influence a customer’s choice of a service provider and impose significant competitive constraints, especially in high technology industries that experience rapid innovation.302 This section presents data in four broad categories reflecting various elements of non-price rivalry among mobile wireless service providers: investment; network coverage and technology upgrades; quality of service; and differentiation in handsets/devices and advertising/marketing. A. Investment 106. Service providers can expand their network coverage and capacity through increased investment in, and expansion of, their existing assets and infrastructure.303 Providers may make such strategic capital expenditure (“CAPEX”) decisions to differentiate their service offerings from those of their rivals by becoming the first to deploy a particular upgrade or new network technology. Over the past five years, wireless providers in the U.S. have made capital investments of approximately $146 billion.304 107. As shown in Chart VI.A.1, wireless providers spent an incremental $32.1 billion in 2014, which is a decline of approximately 3 percent from the $33.1 billion invested in 2013. Based on UBS data, AT&T, Verizon Wireless, Sprint, and T-Mobile spent a combined $13.9 billion in the first half of 2015 and $31.1 billion in 2014, accounting for close to 100 percent of total industry capital investment as tracked by UBS in these time periods. As seen in Chart VI.A.1, AT&T and Verizon Wireless consistently made more capital investments in absolute CAPEX dollars in each quarter than did either Sprint or T-Mobile. However, if calculated as a percentage of service revenues, as of the second quarter of 2015, for example, both Sprint and T-Mobile invested approximately 19 percent of their total service revenues, as compared to approximately 14 percent for AT&T, and approximately 18 percent for Verizon Wireless.305 In addition, one should not place too much emphasis on absolute capital expenditures at any given point in time, as that will not provide the full picture of a service provider’s investment strategy given the cyclical nature of such investments.306 302 See Seventeenth Report, 29 FCC Rcd at 15393 ¶ 168. 303 Importantly, service providers also expand into new geographic areas and/or upgrade networks in existing markets after adding to their spectrum portfolios through participation in spectrum auctions and secondary market transactions, as discussed above. 304 See CTIA Year-End 2014 Wireless Indices Report, Table 35, at p. 98. CTIA’s figure includes incremental investment in currently operational systems, including expenditures for building operating systems, land and capital leases, and all tangible non-system capital investment, but does not include the cost of spectrum licenses purchased at auctions or other acquisition processes or greenfield builds. 305 See UBS US Wireless 411: Version 57, Figure 62. This compares to percentages of approximately 20% for AT&T, approximately 13% for Sprint, approximately 18% for T-Mobile, and approximately 14% for Verizon Wireless, as of the first quarter of 2014. See id. 306 The Sixteenth Report noted that CAPEX in system/network assets may be cyclical or “lumpy” because technological change in the mobile wireless service industry is commercially implemented in successive generations of technologies. Consequently, CAPEX may vary between periods and fluctuations in measures of CAPEX are consistent with the cyclical nature of technological adoption in the mobile wireless service industry. See Sixteenth Report, 28 FCC Rcd at 3842 ¶ 215. Federal Communications Commission DA 15-1487 68 Source: UBS US Wireless 411: Version 55, Figure 54. UBS US Wireless 411: Version 57, Figure 60. T-Mobile and MetroPCS merged in 2013; AT&T acquired Leap in 2014. 108. Looking beyond the quarterly data in Chart VI.A.1, we see the variation in capital expenditures by the four nationwide service providers during the last five and a half years. Chart VI.A.2 below presents annual capital expenditures for the four nationwide service providers from 2010 through the first half of 2015. From 2010 through the end of 2014, AT&T and Verizon Wireless increased their nominal investment (with the exception of a dip in investment in 2012 for Verizon Wireless). Sprint increased its capital investment from 2010 to 2013, but decreased its capital investment in 2014, while T-Mobile’s capital expenditures decreased between 2010 and 2011, and then increased, with a sharp spike from 2012 to 2013. Variations in CAPEX may vary across service providers for several reasons. First, service providers follow different technological migration paths, which may be on different timeframes. Recently, the industry has followed multiple technological migration paths for LTE upgrades, with each service provider implementing its own sequence of upgrades. Second, service providers often base their investment decisions on an assessment of how network deployments and upgrades may affect future earnings. Third, the timing of network investments often has a strategic component vis-à-vis rivals, as noted above. Finally, access to capital may be more constrained for some service providers, and this may require reallocation of their investment.307 307 According to NTCA – The Rural Broadband Association (“NTCA”), which consists exclusively of small, rural service providers, 62% of the rural providers who were surveyed described the process of obtaining financing for their wireless projects as “somewhat difficult” or “very difficult,” while another 5% found it “virtually impossible.” NTCA 2014 Wireless Survey Report, at pp. 3, 10 (Dec. 2014), available at https://www.ntca.org/images/stories/Documents/Advocacy/SurveyReports/2014ntcawirelesssurvey.pdf. 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 C ap E x ( $ i n m il li o n s) Quarters Chart VI.A.1 Capital Expenditure by U.S. Mobile Wireless Service Providers 1Q12 - 2Q15 AT&T VZW Sprint T-Mobile US Cellular Metro PCS Leap Other Providers Federal Communications Commission DA 15-1487 69 Source: UBS US Wireless 411: Version 55, Figure 54. UBS US Wireless 411: Version 57, Figure 60. B. Network Coverage and Technology Upgrades 109. One critical way in which mobile wireless service providers differentiate themselves is through the coverage and speed of their mobile broadband networks. During 2014 and early 2015, as discussed in more detail below, several service providers continued to upgrade and expand their networks with technologies that enable faster data transfer speeds, while other service providers announced plans to make additional upgrades in the near future. We consider in this section current network coverage by service provider, technology and roaming by provider, and future network deployment plans by service provider. 1. Current Coverage by Service Provider 110. This section of the Report presents mobile wireless coverage and LTE broadband coverage. Similar to the analysis of nationwide mobile coverage in section III, our analysis is based on U.S. census blocks overlaid on coverage maps that are provided to the Commission through a contract with Mosaik Solutions. We note that these coverage estimates represent deployment of mobile networks and do not indicate the extent to which service providers actually offer service to any or all residents in the covered areas. While recognizing that this analysis likely results in an overstatement of the coverage experienced by consumers because of limitations in Mosaik data and the coverage calculation methodology,308 we find that this analysis is useful because it provides a general baseline that can be compared over time across network technologies and service providers.309 In addition to the Mosaik data, this Report introduces our initial summary analysis of nationwide mobile coverage based on Form 477 data provided by the wireless service providers themselves. These data will be analyzed using the same 308 The centroid methodology considers a census block as covered if the centroid falls within the boundaries of the census block. By definition, this implies that for regular shaped blocks, at least 50% of the area is covered. This methodology may potentially overstate actual coverage, especially for larger rural blocks. 309 See section III.A. supra. $8,438 $8,973 $8,857 $9,425 $10,515 $5,545 $9,171 $9,764 $10,795 $11,191 $11,383 $3,992 $1,444 $2,416 $4,884 $6,833 $4,886 $2,154 $2,819 $2,729 $2,901 $4,241 $4,317 $2,173 $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $14,000 $16,000 2010 2011 2012 2013 2014 1H2015 C ap it al E x p en d it u re s ($ M il li o n s) Chart VI.A.2 Yearly Capital Expenditures by Service Provider 2010 - 1H2015 Verizon Wireless AT&T Federal Communications Commission DA 15-1487 70 methods described above for the Mosaik data. We note that moving forward, and as also indicated in section III above, we anticipate that the Form 477 data will be our primary source for the analysis of coverage in the mobile wireless marketplace. a. Mobile Wireless Network Coverage by Service Provider 111. Subject to the data limitations just noted, and as set out in more detail in section III. above, mobile wireless coverage by service provider, estimated using Mosaik data, is presented in Chart VI.B.1.310 Each of the four nationwide service providers covered census blocks containing at least approximately 90 percent of the population with their respective mobile wireless networks as of July 2015. In terms of road miles, Chart VI.B.1 shows that Verizon Wireless covered around 87 percent, AT&T covered around 90.5 percent, Sprint covered around 47 percent, and T-Mobile covered around 58 percent as of that date. Verizon Wireless and AT&T each covered over 60 percent of the U.S. land area with their respective mobile wireless networks, while T-Mobile and Sprint each covered less than 35 percent of the U.S. land area. Chart VI.B.2 presents the results from the same analysis using Form 477 data. Both sources indicate, in general, that at least 90 percent of the population is covered by each of the nationwide service providers, but network coverage by land area is much more limited. Chart VI.B.1 Estimated Mobile Wireless Coverage in the U.S. by Service Provider, Mosaik, July 2015 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 310 See Appendix Tables VI.B.i and VI.B.ii for more detailed data on estimated mobile wireless voice coverage and mobile 3G or better coverage by service provider. 34.0% 24.0% 68.4% 63.7% 58.0% 47.3% 90.5% 87.0% 91.6% 89.7% 99.3% 97.3% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint AT&T Verizon Wireless Mobile Wireless Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Source: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 71 Chart VI.B.2 Estimated Mobile Wireless Coverage in the U.S. by Service Provider, Form 477, December 2014 Note: The coverage calculation methodology has certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. b. LTE Broadband Coverage by Service Provider 112. LTE broadband coverage by service provider, estimated using Mosaik data, is presented in Chart VI.B.3.311 Verizon Wireless and AT&T each covered census blocks containing more than 96 percent of the population with LTE as of July 2015, while Sprint and T-Mobile each covered approximately 85 percent. Verizon Wireless covered approximately 85 percent of road miles and 61 percent of the land area, while AT&T covered approximately 72 percent of road miles and 46 percent of the land area. Sprint and T-Mobile each covered approximately 40 percent of U.S. road miles and no more than approximately 20 percent of the U.S. land area with LTE. Chart VI.B.4 presents the results from the same analysis using Form 477 data as of December 31, 2014. According to both Mosaik and Form 477 data, over 80 percent of the population is covered by each of the nationwide service providers, but network coverage by land area is much more limited. 311 See Appendix Table VI.B.iii. 33.6% 25.3% 70.9% 66.7% 56.1% 48.4% 90.4% 86.8% 90.8% 90.5% 99.2% 97.2% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint AT&T Verizon Wireless Mobile Wireless Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e D ec em b er 2 0 1 4 % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Source: Form 477 Data, 2010 Census Federal Communications Commission DA 15-1487 72 Chart VI.B.3 Estimated Mobile LTE Broadband Coverage in the U.S. by Service Provider, Mosaik, July 2015 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Chart VI.B.4 Estimated Mobile LTE Broadband Coverage in the U.S. by Service Provider, Form 477, December 2014 Note: The coverage calculation methodology has certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 20.1% 17.7% 46.4% 60.9% 41.0% 38.6% 72.2% 84.6% 85.2% 85.0% 96.7% 97.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint AT&T Verizon Wireless LTE Broadband Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Source: Mosaik Data, 2010 Census 16.7% 18.5% 46.6% 63.5% 35.7% 39.4% 70.9% 84.3% 82.2% 85.3% 96.0% 96.9% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint ATT Verizon Wireless LTE Broadband Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (D ec em b er 3 0 1 4 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square Miles Source: Form 477 Data, 2010 Census Federal Communications Commission DA 15-1487 73 c. Urban/Rural Comparisons (i) Mobile Wireless Network Coverage 113. A comparison of rural and non-rural coverage by service provider, estimated using Mosaik data, is presented in Chart VI.B.5.312 Each of the four nationwide service providers covered census blocks containing more than 94 percent of the non-rural population with their mobile wireless voice networks, as of July 2015.313 In terms of the rural population, Verizon Wireless and AT&T covered census blocks containing approximately 91 percent, respectively, while Sprint and T-Mobile covered census blocks containing approximately 57 and 64 percent, respectively. Each of the four nationwide service providers covered a significantly higher percentage of non-rural than rural land area and road miles. The results based on the Form 477 data are similar. Chart VI.B.5 Estimated Mobile Wireless Voice Coverage in Non-Rural and Rural Areas by Provider; July 2015 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 312 As noted above, while the Communications Act does not include a statutory definition of what constitutes a rural area, since its 2004 Report and Order concerning the deployment of wireless services in less populated areas, the Commission has used a “baseline” definition of rural as a county with a population density of less than 100 people per square mile. See Facilitating the Provision of Spectrum-Based Services to Rural Areas and Promoting Opportunities for Rural Telephone Companies To Provide Spectrum-Based Services, WT Docket 02-381, Report and Order, 19 FCC Rcd 19078, 19087-88 (2004). 313 See Appendix Tables VI.B.iv through VI.B.vii for more detailed data on estimated mobile wireless voice coverage and mobile 3G or better coverage by provider in rural and non-rural areas, respectively. 63.9% 56.8% 91.4% 90.8% 97.3% 94.6% 99.8% 98.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint AT&T Verizon Wireless Mobile Wireless Voice Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of Non-Rural U.S. Population % of Rural U.S. Population Source: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 74 (ii) LTE Broadband Coverage 114. As seen in Chart VI.B.6, each of the four nationwide providers covered census blocks containing at least 95 percent of the non-rural population with their LTE networks, as of July 2015.314 In terms of the rural population, Verizon Wireless and AT&T covered census blocks containing approximately 92 percent and 85 percent, respectively, while Sprint and T-Mobile covered census blocks containing approximately 48 and 47 percent of the rural population, respectively. Each of the four nationwide providers covered a higher percentage of non-rural than rural land area and road miles. There are still significant amounts of the rural population in the United States without LTE broadband coverage, as compared to the urban population. The results based on the Form 477 data are similar. Chart VI.B.6 Estimated Mobile LTE Coverage in Non-Rural and Rural Areas by Service Provider, July 2015 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 2. Service Provider Network Deployments 115. Network investment remains a centerpiece of service providers’ efforts to improve their customers’ mobile wireless service experience. Over the past several years, service providers have upgraded and expanded their networks with technologies that enable faster data transfer speeds, and all of the major mobile wireless providers have now deployed LTE to a significant extent. The following section includes a discussion of the deployment strategies and mobile network upgrades for each of the top five service providers. 314 See Appendix Tables VII.B.viii and VII.B.ix for more detailed data on estimated LTE broadband coverage by service provider in rural areas and non-rural areas, respectively. 47.3% 48.1% 85.1% 92.1% 94.0% 93.6% 99.4% 98.2% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% T-Mobile Sprint AT&T Verizon Wireless LTE Broadband Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of Non-Rural U.S. Population % of Rural U.S. Population Source: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 75 Table VI.B.1 LTE Deployment Reported by Selected Mobile Wireless Service Providers, October 2015 Service Provider LTE Deployment as of October 2015 Verizon Wireless Verizon Wireless covers 308 million people in over 500 US markets with LTE, and over 400 of the LTE markets are served by XLTE.315 Voice and Non-4G LTE data services are available outside their 4G LTE coverage area.316 AT&T AT&T covers over 320 million people with its voice and data service, with over 308 million people covered by its 4G LTE network.317 Sprint Sprint’s 4G LTE network reaches more than 280 million people.318 Sprint Spark (enhanced LTE) is available in limited markets, and on devices with enabled tri-band LTE capability.319 T-Mobile T-Mobile's 4G LTE network will reach over 300 million people by year-end 2015.320 It has launched its Extended Range LTE, which it claims reaches 2 times farther and works 4 times better in buildings compared to traditional LTE.321 116. Verizon Wireless: Verizon Wireless’s LTE network covered approximately 308 million people as of October 2015, as seen in Table VI.B.1 above. Verizon Wireless continued to augment its Upper 700 MHz C Block LTE network with AWS-1 spectrum, and has now launched XLTE in more than 400 markets.322 The company has also begun to re-farm PCS spectrum to use for LTE.323 This type of carrier aggregation, along with 315 See “The Verizon Wireless 4G LTE Network,” available at http://www.verizonwireless.com/news/LTE/Overview.html; “Our Technology: Network,” available at https://www.verizonwireless.com/aboutus/technology/network.html. 316 See “4G LTE,” available at http://www.verizonwireless.com/landingpages/4g-lte/. 317 See “About Our Network,” available at http://about.att.com/news/wireless-network.html; “About AT&T,” available at https://www.att.com/network/en/index.html; “1Q 2015 AT&T,” available at http://www.att.com/Common/about_us/pdf/att_btn.pdf. 318 See “Sprint Network Information Center,” available at http://newsroom.sprint.com/presskits/sprint-network-vision- information-center.htm. 319 As noted above, the Sprint WiMax network was shut down in November 2015. See “Sprint to Shutter WiMAX Network Around Nov. 6, 2015,” available at http://www.fiercewireless.com/story/sprint-shutter-wimax-network-around-nov-6- 2015/2014-10-07. 320 See “Coverage,” available at http://www.t-mobile.com/coverage.html. See also “T-Mobile Doubles 4G Coverage to Take on AT&T and Verizon.” available at http://www.cnet.com/news/t-mobile-doubles-4g-coverage-to-take-on-at-t-and-verizon/; “We’ve Doubled Our 4G LTE Coverage in the Last Year,” available at https://explore.t-mobile.com/4G-lte-network. 321 See “Coverage,” available at http://www.t-mobile.com/coverage.html; “T-Mobile Adds 2.3 Million New Customers in Q3 2015,” available at http://www.tmonews.com/2015/10/t-mobile-adds-2-3-million-new-customers-in-q3-2015/; “We’ve Doubled Our 4G LTE Coverage in the Last Year,” available https://explore.t-mobile.com/4G-lte-network. 322 See “Customers Celebrate One Year of XLTE,” available at http://www.verizonwireless.com/news/article/2015/05/customers-celebrate-one-year-of-xlte.html. 323 See “Verizon to Launch Carrier Aggregation, More LTE Advanced Features in 2015,” available at http://www.fiercewireless.com/story/verizon-launch-carrier-aggregation-more-lte-advanced-features-2015/2014-12-04. “In addition to New York City, Verizon has started refarming PCS spectrum for LTE in Cleveland and around 10 other markets.” Federal Communications Commission DA 15-1487 76 a heterogeneous network of macrocells and small cells, will be combined with other network improvements to implement LTE advanced.324 In addition, Verizon Wireless has started to roll out VoLTE service nationwide.325 In August 2015, Verizon Wireless announced that it was aggressively working on VoLTE transition and already had 4 million customers using the service.326 Moving forward, Verizon Wireless plans to deploy LTE-Unlicensed technology in 2016,327 and in order to promote this technology, Verizon Wireless formed the LTE-U forum in 2014, in cooperation with Alcatel-Lucent, Ericsson, and Qualcomm Technologies, Inc., and Samsung.328 117. AT&T: AT&T’s LTE network covered over 308 million people as of October 2015, as seen in Table VI.B.1 above. AT&T continues to focus on DAS and small cells,329 and to deploy commercial mobile broadband services using carrier aggregation technology – which is part of the LTE Advanced specifications – to combine transmissions across either AWS-1 or PCS spectrum with Lower 700 MHz B and C Block spectrum.330 In addition, AT&T began deploying LTE on its 2.3 GHz WCS spectrum in the summer of 2015.331 AT&T continues to launch HD Voice and Voice over LTE (“VoLTE”) in select markets, and has stated that VoLTE is in fact necessary to support voice roaming for CDMA service providers.332 Further, AT&T will trial LTE- Unlicensed technology in late 2015 or early 2016, but will wait to deploy it until the Wi-Fi sharing rules are more defined.333 118. Sprint: Sprint’s 4G LTE service was available to approximately 280 million Americans in October 2015, as shown in Table VI.B.1, above. In its comments, Sprint notes the importance of its Network Vision project, which is an initiative to consolidate Sprint’s networks and technologies into a single nationwide See also “Oppenheimer 18th Annual Technology, Internet and Communications Conference,” at 3, available at http://www.verizon.com/about/investors/oppenheimer-18th-annual-technology-internet-communications-conference, Aug. 11, 2015. 324 See “The Verizon Wireless 4G LTE Network,” available at http://www.verizonwireless.com/news/LTE/Overview.html. 325 See “Introducing HD Voice and Video Calling,” available at http://www.verizonwireless.com/wcms/consumer/advanced- calling.html; http://www.verizonwireless.com/support/advanced-calling-faqs/. 326 See “Verizon Aggressively Working on VoLTE Transition,” available at http://www.lightreading.com/mobile/volte-rich- communications/verizon-aggressively-working-on-volte-transition-/d/d-id/717538. 327 See “Verizon’s LTE-U Forum Completes LTE Unlicensed Tests on Way Toward Carrier’s 2016 Deployment in 5 GHz, 3.5 GHz,” available at http://www.fiercewireless.com/tech/story/verizons-lte-u-forum-completes-lte-unlicensed-tests-way- toward-carriers-201/2015-03-03. 328 See LTE-U Forum website, available at http://www.lteuforum.org/. 329 See William Ho, “As T-Mobile and Sprint Catch Up to Verizon and AT&T on LTE Coverage, Capacity Comes into Focus,” available at http://www.fiercewireless.com/story/hos-perspective-t-mobile-and-sprint-catch-verizon-and-att-lte- coverage-capa/2015-05-19. 330 See “AT&T’s Keathley Shares a 5G Vision that Includes LTE-A, Wi-Fi, SON and More,” available at http://www.fiercewireless.com/tech/story/atts-keathley-shares-5g-vision-includes-lte-wi-fi-son-and-more/2015-01-21; William Ho, “As T-Mobile and Sprint Catch Up to Verizon and AT&T on LTE Coverage, Capacity Comes into Focus,” available at http://www.fiercewireless.com/story/hos-perspective-t-mobile-and-sprint-catch-verizon-and-att-lte-coverage- capa/2015-05-19. 331 See “AT&T Begins Deploying 2.3 GHz WCS Spectrum for LTE,” available at http://www.fiercewireless.com/story/att- begins-deploying-23-ghz-wcs-spectrum-lte/2015-09-09. 332 See FCC website, AT&T April 3, 2015 Ex Parte, Promoting Interoperability in the 700 MHz Commercial Spectrum, WT Docket No. 12-69, available at http://apps.fcc.gov/ecfs/document/view?id=60001042267&cmpid=horizontalcontent; AT&T website, “AT&T HD Voice,” available at http://www.att.com/shop/wireless/features/hd-voice.html. 333 See “AT&T to Trial LTE-U by Early 2016,” available at http://www.fiercewireless.com/tech/story/att-trial-lte-u-early- 2016/2015-09-09. Federal Communications Commission DA 15-1487 77 3G and 4G network.334 In addition, Sprint has been promoting Wi-Fi calling on its network.335 In June 2015, Sprint indicated as part of its network densification plan, the company also planned on deploying VoLTE technology.336 119. T-Mobile: T-Mobile’s LTE coverage has just now reached over 300 million people.337 T-Mobile has also continued to launch 4G LTE in a 15×15 megahertz deployment in certain markets.338 In addition, T- Mobile now offers Wi-Fi calling.339 T-Mobile has launched VoLTE across its 4G LTE network.340 T-Mobile plans to launch LTE-Unlicensed in the coming year.341 120. U.S. Cellular: In March of 2015, U.S. Cellular announced that it plans to add over 600 4G LTE cell sites in 2015, providing access to 4G LTE to 98 percent of its approximately 4.8 million customers.342 Currently, nearly 90 percent of customers have access to 4G LTE in U.S. Cellular network coverage areas.343 In addition, U.S. Cellular is testing VoLTE but has not announced a timeline for its commercial launch.344 3. FCC Initiatives Related to Coverage, Technology, and Roaming 121. When competing mobile wireless service providers deploy compatible network technologies, greater economies of scale in the production of both end-user devices and network infrastructure equipment can result, lowering the unit cost of handsets, chipsets, and other network equipment. This, in turn, may promote more rapid adoption of mobile wireless services, a greater variety of handsets, and more price competition. On October 25, 2013, the Commission adopted the 700 MHz Interoperability Report and Order and Order of 334 See Sprint Comments at 5. 335 See Sprint Comments at 5-6. See also “FAQs About Wi-Fi Calling,” available at http://support.sprint.com/support/article/FAQs-about-Wi-Fi-Calling/173e331f-8423-453e-93cb-4688f6a91f67#!/; “Sprint iPhone Customers Take Full Advantage of Wi-Fi Calling with Apple’s Latest Software Update,” available at http://investors.sprint.com/news--investor-events/newsroom/press-release-details/2015/Sprint-iPhone-Customers-Take-Full- Advantage-of-Wi-Fi-Calling-with-Apples-Latest-Software-Update/default.aspx. 336 See “Sprint Gets OK From SoftBank To Move Ahead On Network Densification Plan,” available at http://www.fiercewireless.com/story/sprint-gets-ok-softbank-move-ahead-network-densification-plan/2015-06-03. 337 See “T-Mobile US Inc. at Morgan Stanley European Technology, Media and Telecom Conference,” Thomason Reuters, at 6 (Nov. 12, 2015) (“We just crested over 300 million covered people in the U.S. with our LTE network. We are just north of 302 million . . . A third of our traffic now is on VoLTE, so voiceover LTE.” Also, at 9 “We currently sell in about 230 million POPs so that’s our effective selling footprint . . . our organic network covers over 300 now.”) 338 See “Why 20×20 4G LTE is Awesome,” available at https://support.t-mobile.com/community/coverage/blog. 339 See “Now Every Wi-Fi Connection Works Like a T-Mobile Tower,” available at http://www.t-mobile.com/offer/wifi- calling-wifi-extenders.html. 340 See “Firing on All Cylinders,” available at http://newsroom.t-mobile.com/issues-insights-blog/firing-on-all-cylinders- earnings-jdp.htm. 341 See “Verizon, T-Mobile Push Unlicensed LTE Forward – But Concerns Remain,” available at http://www.fiercewireless.com/story/verizon-t-mobile-push-unlicensed-lte-forward-concerns-remain/2015-03-03; “T-Mobile Will Launch LTE in the Wi-Fi Airwaves in 2016,” available at https://gigaom.com/2015/03/01/t-mobile-will-launch-lte-in- the-wi-fi-airwaves-in-2016/. 342 See “U.S. Cellular Announces New Markets To Receive 4G LTE Service in 2015,” available at http://www.uscellular.com/about/press-room/2015/USCELLULAR-ANNOUNCES-NEW-MARKETS-TO-RECEIVE-4G- LTE-SERVICE-IN-2015.html. 343 See “Press Room FAQs,” available at http://www.uscellular.com/uscellular/support/faq/faqDetails.jsp?topic=press- room.html. 344 See “U.S. Cellular is Testing VoLTE in 3 Markets But Has No Timeline For Commercial Launch,” available at http://www.fiercewireless.com/story/us-cellular-testing-volte-3-markets-has-no-timeline-commercial-launch/2015-11-10. Federal Communications Commission DA 15-1487 78 Proposed Modification to implement a voluntary industry solution that is designed to establish interoperable LTE devices in the Lower 700 MHz band.345 Since the Order was adopted, service providers have progressively deployed the Lower 700 MHz A Block spectrum and are continuing to offer more Band Class 12 devices, facilitating the creation of a robust Band Class 12 devices ecosystem.346 In addition, the Commission has also adopted specific interoperability requirements for the AWS-3 band, as well as for the 600 MHz Band.347 122. In order to encourage mobile network deployment into unserved or underserved areas, the Commission adopted rules creating the Mobility Fund in November 2011.348 The Mobility Fund uses Universal Service Fund reserves to support the deployment of current- or future-generation mobile network technologies that provide mobile voice and Internet services.349 For Mobility Fund Phase I, the Commission is providing up to $300 million in one-time support payments, plus up to $50 million dedicated to Tribal lands, which were awarded through reverse auctions.350 The Commission is currently exploring whether to retarget Mobility Fund Phase II ongoing support to ensure the continued deployment and preservation of LTE mobile broadband service and preservation of mobile voice and broadband service in areas that otherwise would not have such service through marketplace forces.351 123. No facilities-based service provider – including the four nationwide service providers – has built out its entire licensed service area, and consequently all employ roaming to some extent to fill gaps in their 345 See Promoting Interoperability in the 700 MHz Commercial Spectrum, Requests for Waiver and Extension of Lower 700 MHz Band Interim Construction Benchmark Deadlines, WT Docket Nos. 12-69, 12-332, Report and Order and Order of Proposed Modification, 28 FCC Rcd 15122 (2013) (“700 MHz Interoperability Report and Order and Order of Proposed Modification”). 346 Band Class 12 devices support the Lower 700 MHz spectrum A, B, and C blocks whereas Band Class 17 devices only support the Lower 700 MHz spectrum B and C blocks. T-Mobile has continued to deploy LTE in “141 market areas”, through the second quarter of 2015, using its Lower 700 MHz A Block spectrum, while USCC has also deployed LTE using its Lower 700 MHz spectrum in “over 100 markets.” In addition, AT&T states that it has “fully deployed” the Multiple Frequency Band Indicator (“MFBI”) feature to support both Band Class 12 and Band Class 17 devices throughout its network, and will start transitioning to Band Class 12 devices “late in 2015 and/or early 2016.” AT&T also maintains that it is continuing discussions with several Band Class 12 service providers to offer LTE roaming service on its network. See AT&T’s Interoperability Status Report, Promoting Interoperability in the 700 MHz Commercial Spectrum, WT Docket No. 12-69, September 14, 2015, Attachment 1. 347 See AWS-3 Report and Order, 29 FCC Rcd at 4694-4700 ¶¶ 225-31; Incentive Auctions Report and Order, 29 FCC Rcd at 6866-69 ¶¶ 731-37. 348 See Connect America Fund; A National Broadband Plan for Our Future; Establishing Just and Reasonable Rates for Local Exchange Carriers; High-Cost Universal Service Support, Developing an Unified Intercarrier Compensation Regime; Federal- State Joint Board on Universal Service; Lifeline and Link-Up; Universal Service Reform – Mobility Fund, WC Docket No. 10- 90, et al., Report and Order and Further Notice of Proposed Rulemaking, 26 FCC Rcd 17663 (2011). 349 See id. 350 See Mobility Fund Phase I Auction Closes; Winning Bidders Announced for Auction 901, AU Docket No. 12-25, Public Notice, 27 FCC Rcd 12031 (WTB 2012) (“Auction 901 Closing Public Notice”). Mobility Fund Phase I disbursements were authorized beginning April 2013 and are anticipated to continue through 2016. See Mobility Fund Phase I Support Authorized for Seven Winning Bidders; Defaults on Two Auction 901 Winning Bids Determined, AU Docket No. 12-25, Public Notice, 28 FCC Rcd 5599 (WTB, WCB 2013); Tribal Mobility Fund Phase I Auction Closes Winning Bidders Announced for Auction 902, Public Notice, 29 FCC Rcd 1974 (WTB 2014). 351 See Connect America Fund; Universal Service Reform – Mobility Fund; ETC Annual Reports and Certifications; Establishing Just and Reasonable Rates for Local Exchange Carriers; Developing an Unified Intercarrier Compensation Regime, WC Docket No. 10-90, et al., Report and Order, Declaratory Ruling, Order, Memorandum and Order, Seventh Order on Reconsideration, and Further Notice of Proposed Rulemaking, 29 FCC Rcd 7051 (2014). Federal Communications Commission DA 15-1487 79 coverage.352 Many non-nationwide service providers are able to offer their customers coverage that is national in scope through roaming agreements with other mobile wireless service providers, and accordingly, roaming remains particularly important for small and regional service providers.353 In recent years, the Commission has taken actions to facilitate roaming arrangements: In 2007, for instance, it clarified that automatic voice roaming is a common carrier obligation for CMRS providers.354 In April 2010, the Commission adopted the Roaming Order on Reconsideration, which eliminated the home roaming exclusion and established the same general obligation to provide automatic voice roaming, regardless of whether the provider requesting roaming holds spectrum in an area.355 In April 2011, the Commission issued the Data Roaming Order,356 which requires facilities-based service providers of commercial mobile data services, whether or not such providers also offer CMRS, to offer data roaming arrangements to other mobile data service providers on commercially reasonable terms and conditions, subject to certain limitations.357 124. On May 27, 2014, T-Mobile USA, Inc. filed a petition for expedited declaratory ruling in which it asked the Commission to provide guidance for determining whether the terms of a data roaming agreement meet the “commercially reasonable” standard adopted in the Data Roaming Order.358 The petition proposed four benchmarks for assessing commercial reasonableness: (i) retail rates; (ii) international roaming rates; (iii) MVNO/resale rates; and (iv) other domestic roaming rates. The petition also sought guidance on the extent and nature of service providers’ build-out as a factor in determining commercial reasonableness and the presumption of reasonableness for existing data roaming agreements. The Wireless Telecommunications Bureau released an order on December 18, 2014, granting the T-Mobile petition.359 The Order found that the data roaming rule was 352 See Seventeenth Report, 29 FCC Rcd at 15399 ¶ 180. 353 See id.; see also CCA Comments at 18-19; NTCA Comments at 3-4 (arguing that regional and local service providers offer a small footprint and need to partner with other service providers through roaming agreement to offer their subscribers competitive expanded coverage). By definition, MVNOs and resellers rely on using the networks of one or several facilities-based service providers to compete. Depending on the particular arrangement, customers of MVNOs and resellers may have limited or no access to networks other than that of the underlying wholesale provider. For example, the coverage experienced by customers of TracFone’s Straight Talk varies depending on the underlying wholesale provider. A Straight Talk customer obtaining service on an underlying CDMA network, for instance, will not have access to voice and SMS service when outside the coverage of that network, but a Straight Talk customer whose service is provided on an underlying GSM network, by contrast, may have voice and SMS service when outside that network’s coverage area. See “About Straight Talk,” available at http://www.straighttalk.com/wps/portal/home/h/about; “Terms and Conditions: Straight Talk Wireless,” available at http://www.straighttalk.com/wps/portal/home/h/legal/terms-and-conditions. 354 See Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers, WT Docket No. 05-265, Report and Order and Further Notice of Proposed Rulemaking, 22 FCC Rcd 15817, 15828 ¶ 27 (2007) (“2007 Roaming Order and FNPRM”) (“[W]e recognize that automatic roaming benefits mobile telephony subscribers by promoting seamless CMRS service around the country, and reducing inconsistent coverage and service qualities.”) 355 See Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers and Other Providers of Mobile Data Services, WT Docket No. 05-265, Order on Reconsideration and Second Further Notice of Proposed Rulemaking, 25 FCC Rcd 4181, 4182 ¶ 2 (2010) (“Roaming Order on Reconsideration”). 356 See Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers and Other Providers of Mobile Data Services, WT Docket No. 05-265, Second Report and Order, 26 FCC Rcd 5411 (2011) (“Data Roaming Order”). 357 See id., 26 FCC Rcd at 5418-28 ¶¶ 13-31. 358 See Wireless Telecommunications Bureau Seeks Comment on Petition for Expedited Declaratory Ruling by T-Mobile USA, Inc. Regarding Data Roaming Obligations, WT Docket No. 05-265, Public Notice, 29 FCC Rcd 6035 (WTB 2014). 359 See Reexamination of Roaming Obligations of Commercial Mobile Radio Service Providers and Other Providers of Mobile Data Services, WT Docket No. 05-265, Declaratory Ruling, 29 FCC Rcd 15483 (WTB 2014) (“Data Roaming Federal Communications Commission DA 15-1487 80 intended to permit consideration of the totality of the facts and therefore to permit a complaining party to adduce evidence as to whether proffered roaming rates are substantially in excess of retail rates, international rates, and MVNO/resale rates, as well as a comparison of proffered roaming rates to domestic roaming rates as charged by other service providers.360 This Order noted that the probative value of these other rates as reference points will depend on the facts and circumstances of any particular case, including all of the factors set forth in the Data Roaming Order. In addition, the Order provided guidance with respect to the presumption regarding existing agreements and consideration of the build-out factor in determining commercial reasonableness.361 We note that since then, a number of confidential staff-level mediations have occurred, and certain complaints were filed and remain pending. C. Quality of Service 125. Key characteristics for mobile wireless performance include network speeds, latency, and packet loss. The Commission has recognized the importance of accurate and timely data on these characteristics. This Report will primarily analyze speed data using the Ookla Net Index data, data from the FCC Speed Test App, RootMetrics data, and the CalSPEED drive-test data gathered by the California Public Utility Commission (“CPUC”), subject to the limitations described below. 126. Mobile broadband network service quality experienced by consumers may vary greatly with a number of real world factors such as the service provider’s received signal quality, cell traffic loading and network capacity in different locations, as well as the capability of consumers’ devices.362 First, mobile connection quality will vary based on the location of the receiving device in reference to the transmitting device, which is often a cellular tower. If the receiving device (and the person using it) is behind a wall, blocked by terrain or otherwise has an impaired connection with the tower, service will be degraded or not available. Second, the performance of the broadband connection degrades over distance to the tower, even with a clear line of sight. Performance at the edge of a tower’s coverage is not equal to performance close to the tower. Third, cellular signals are shared by many users – the more simultaneous usage, the lower the potential performance of any one connection.363 It is also important to note that for all mobile technologies, speed and performance measurements are only valid when a wireless connection can be accessed. “Dead zones” and loss of signal reduce wireless effectiveness.364 Moreover, from the customer’s perspective, overall network performance is the product of more than network quality alone and often reflects differences in device capability as well.365 For data services, network quality as perceived by the customer may also be use-, case-, or application-dependent (e.g., a consumer Declaratory Ruling”). We note that AT&T and Verizon Wireless filed applications for review of the declaratory ruling, and those applications for review remain pending. 360 See id., 29 FCC Rcd at 15486 ¶ 9. 361 See id., 29 FCC Rcd at 15491, 15492 ¶¶ 25, 28. 362 For example, the received signal quality is dependent on the service provider’s deployed cell site density, low/high frequency radio wave propagation losses, user locations, indoor obstructions and outdoor foliage or clutter, weather, inter-cell interference conditions, and wireless network optimization parameters. The cell traffic loading or demand is dependent on the overall number of concurrent active mobile broadband users sharing the same cell, which in turn depends on user locations, the day of the week, and the time of the day. The capacity of a service provider’s wireless network is dependent on the deployed mobile wireless technology, sites and equipment, available bandwidth, and enhanced backhaul connections. See Sixteenth Report, 28 FCC Rcd at 3894 ¶ 293. 363 See The FCC Omnibus Broadband Initiative (“OBI”); Broadband Performance: OBI Technical Paper No. 4, at p. 19, available at https://apps.fcc.gov/edocs_public/attachmatch/DOC-300902A1.pdf. 364 See id. at pp. 19-20. 365 The capability of consumer devices (e.g., smartphones, tablets, USB dongles, and laptops) could result in users experiencing different data speeds on the same mobile wireless broadband network. Even differing capabilities within each device category, such as smartphone processing power and memory, could result in better user experiences on 4G networks. Federal Communications Commission DA 15-1487 81 who solely uses e-mail may view the quality of the network differently than one who streams video regularly). Further, consumers may place more weight on one particular aspect of network quality than another – such as coverage or peak data speeds – when choosing their mobile wireless services. 127. In recognition of the effects of these different parameters on mobile network performance, various methodologies are used in evaluating mobile network speeds. The two most prevalent rely on crowdsourced data or structured sample data. Crowdsourced data are user-generated data produced by consumers who voluntarily download speed test applications on their mobile devices.366 Generally, crowdsourced data can bring the benefits of generating a large volume of data at a very low cost and of measuring actual consumer experience on a network in a wide variety of locations, indoor and outdoor. We note, however, that crowdsourced data are often not collected pursuant to statistical sampling techniques, and may require adjustments to construct a representative sample from the raw data. For instance, crowdsourced mobile data come from a self-selected group of users, and there often is little control regarding such parameters as when people implement the test, whether the test is performed indoors or outdoors, the geographic location of the tester, and the vintage of the consumer’s device.367 These issues can be reduced by creating default settings that run the test at random times, and not only when the consumer initiates the test. Structured sample data, by contrast, are generated from tests that control for the location and time of the tests as well as for the devices. Structured sample data may be collected using stationary indoor or outdoor tests, or drive tests. However, these tests are more expensive to conduct, involve significant judgment about when and where the tests are run, often do not involve significant testing indoors or in many rural areas, and typically produce datasets that are not as rich as crowdsourced data – all of which are likely to have some effects on reported results. 128. Ookla is one of the most prominent providers of crowdsourced data.368 The Commission has also made available a mobile app that gathers such data. CalSPEED measures mobile network speeds in California based primarily on drive tests. RootMetrics publishes broadband performance metrics that is largely based on drive test data in 125 U.S. metropolitan areas and in 50 airports, but also incorporates results of some crowdsourced data.369 We discuss our analysis of all four speed metrics in the section below.370 366 These apps commonly collect data on the provider, location of device, download and upload speeds, latency, and packet loss, which are then transmitted to the company or entity that developed the app. In some cases, the apps automatically schedule these tests to run at certain times during the day, while in others, the user has to choose to run the tests. 367 By contrast, crowdsourced fixed broadband speed data, such as those collected by the Commission through SamKnows, can be gathered with more control. The SamKnows whiteboxes are able to measure actual fixed network speed and are not dependent on the vintage of the client hardware or software. In addition, the testers are chosen according to a valid sampling technique. 368 In the Seventeenth Report, PC Magazine’s speed test results were also reported. However, in 2015, PC Magazine used Ookla data to compare download and upload speeds across service providers and so is no longer an independent source. See “How Fast Is Your Internet Connection – Really?” available at http://www.pcmag.com/speedtest. 369 See RootMetrics website, available at http://www.rootmetrics.com/us. 370 In addition to the four speed metrics discussed, speed measurements are also performed by other entities such as by OpenSignal and M-Lab. OpenSignal gathers crowdsourced mobile speed data through the use of their mobile app. This app is available free of charge to Android and iOs users, and it is designed to collect data about cell phone towers and cell phone signal strength. OpenSignal uses these data to generate a publicly available interactive map. See “OpenSignal for Android,” available at http://opensignal.com/android/; “OpenSignal for iPhone,” available at http://opensignal.com/iphone/. The data can also be downloaded using OpenSignal’s free NetworkStats API. See OpenSignal Developer website, available at http://developer.opensignal.com/. Another open source application for measuring network performance on mobile platforms is MobiPerf by M-Lab. See MobiPerf website, available at https://sites.google.com/site/mobiperfdev/. This application is available for Android phones only, and data collected via the application are used to generate a publicly available interactive map. See Open Mobile data Federal Communications Commission DA 15-1487 82 1. Ookla 129. Ookla gathers crowdsourced mobile speed data through the use of their Speedtest mobile app.371 This app is available free of charge to smart phone users, and is designed to test the performance of mobile cellular connections. Once the app is downloaded, the user can periodically measure the speed of their wireless connection. These data are then used to produce Ookla’s Net Index dataset. Because the speed tests rely on the phone’s connection to the server, factors such as congestion, location of the server, proximity and access to a cell tower, and phone quality can affect the result. As presented below and in Appendix VI., the Ookla data show significant variation in different geographies, as well as among service providers. In this Report, we present the nationwide mean and median download (and upload – see Appendix VI.) speeds by service provider. Table VI.C.1 presents Ookla’s nationwide median LTE download speed measurements for the first half of 2015,372 and shows that download speeds vary by service provider.373 The trends in download speeds from 2013 through the first half of 2015 are presented below in Table VI.C.2.374 Table VI.C.1 Ookla Speed Test - Estimated LTE Download Speeds by Service Provider, Nationwide375 Service Provider LTE Download Speeds 1H2015 Mean (Mbps) Median (Mbps) Number of tests (’000s) Verizon Wireless 16.67 13.23 2,867 AT&T 13.27 9.93 3,101 Sprint 10.00 7.73 2,130 T-Mobile 17.31 14.39 4,150 Total 11.20 9.51 12,495 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. map, available at http://openmobiledata.appspot.com/visualization. Anonymized data are also available for download. See MobiPerf, “For researchers,” available at https://sites.google.com/site/mobiperfdev/for-researchers. 371 See Speedtest mobile app, available at http://www.speedtest.net/mobile/. 372 More details can be found in Appendix Tables VI.C.i – VI.C.viii. These tables include data from the 2nd half of 2014, as well as the entire time period including the 2nd half of 2014 and the first half of 2015. 373 One factor that may lead to speed differences between wireless service providers is the composition of currently used smartphones. The Ookla Speedtest application is available for download on iOS, Android, or Windows Phones. Each of these operating systems has evolved over time. Vintage smartphones, which do not support 4G, or possibly even 3G service, are still in use. Based on current and past promotions and partnerships, each wireless service provider may have a customer base with a different smartphone profile, which can directly affect speed measurements. 374 The upload and download speeds for 2013 and the first half of 2014 were estimated using publicly available data, which were previously available at Ookla Net Index, http://www.netindex.com/. The Ookla dataset used for 2013 and the first half of 2014 speeds is based on daily mobile download and upload speed data by city and service provider for 2013 – June 2014. Ookla’s daily, city-level observations can consist of thousands of speed measurements, averaged into one data point, therefore the median speed is actually a median of daily averages. We dropped cities that did not have sufficient observations as well as outliers. The upload and download speeds for the second half of 2014 and the first half of 2015 were provided to the Commission by Ookla. 375 Internet performance data, © 2015 Ookla, LLC. All rights reserved. Published with permission of Ookla. Total mean and median download speeds are calculated using aggregated regional data for the four nationwide providers, as well as any other wireless service providers that appear in the Ookla dataset. Only tests with identifiable regions (states) are included. In addition to the the current table, this is applicable to Table VI .C.2 and Appendix Tables VI.C.i – VI.C.viii. Federal Communications Commission DA 15-1487 83 Table VI.C.2 Ookla Median Download Speed by Service Provider, 2013 – 1H2015 Service Provider Median Download Speed (Mbps) 2013 1H2014 2H2014 1H2015 Verizon Wireless 8.99 10.74 11.86 12.65 AT&T 9.04 9.79 7.29 7.86 Sprint 1.92 2.90 4.68 5.35 T-Mobile 6.16 9.89 10.23 11.04 Source: Net Index data by Ookla, 2013, 1H2014. Ookla Internet performance data, 2H2014, 1H2015. © 2015 Ookla, LLC. All rights reserved. 2. FCC 130. In September 2012, the Commission announced that it was expanding its Measuring Broadband America program to include information on mobile broadband service performance in the United States using a crowdsourced approach.376 The program uses the FCC Speed Test app for Android and iPhone devices to test the speed and performance of volunteers’ smartphone mobile broadband services.377 The FCC Speed Test app is available free of charge for Android phones and for the iPhone.378 The FCC Speed Test app provides the benefits, and has the limitations, described above for crowdsourced mobile data.379 Nationwide median LTE download speed measurements for the first half of 2015, estimated using data collected with the FCC Speed Test App, are presented in Table VI.C.3 and Table VI.C.4 below. The first presents separately data associated with 376 See FCC website, “Measuring Mobile Broadband Performance,” available at http://www.fcc.gov/measuring-broadband- america/mobile. 377 See FCC Speed Test, available at https://play.google.com/store/apps/details?id=com.samknows.fcc&hl=en; FCC Speed Test App, available at https://itunes.apple.com/us/app/fcc-speed-test/id794322383?mt=8. The data collected include speed, latency, and packet loss for both upload and download. 378 The FCC speed test can be set to run automatically in the background on Android phones, but iPhone users must execute the speed test manually. This app allows users to measure their mobile broadband performance and voluntarily report these data to the FCC. Collected data include upload and download speed, latency and packet loss, as well as the wireless performance characteristics of the broadband connection and the kind of handsets and versions of operating systems tested. Several other passive metrics are also recorded, including signal strength of the connection, and device manufacturer and model. See FCC website, “Measuring Mobile Broadband Performance,” available at http://www.fcc.gov/measuring- broadband-america/mobile. 379 The results reported here do not incorporate all attempted tests reported to the Commission. Each individual test report includes information on whether the test completed successfully, whether it timed out due to connection problems, and whether any data from the test are missing. Missing data from a test reflect issues in the operation of the app for that particular test, so that particular test observation is dropped from the dataset. If the test timed out, we replaced the reported speed with a value of zero, under the assumption that no meaningful connection was made. In addition, the top 1% of download and upload speed observations were dropped, to account for outliers. Tests where the user was on a Wi-Fi network were also filtered out. Based on the remaining data, we calculated the overall mean and median download and upload speeds by service provider. We present additional results in Appendix Tables VI.C.ix – VI.C.xx. These tables include data from the second half of 2014, as well as the entire time period including the second half of 2014 and the first half of 2015. Federal Communications Commission DA 15-1487 84 the four nationwide service providers’ flagship brands, while the second includes data we have from all brands offered by the four nationwide service providers, as we presented in the Seventeenth Report.380 Table VI.C.3 FCC Speed Test - Estimated LTE Download Speeds by a Service Provider’s Flagship Brand, Nationwide381 Service Provider 1H2015 Mean LTE Download Speed (Mbps) Median LTE Download Speed (Mbps) Number of tests Verizon Wireless 16.63 11.66 47,114 AT&T 11.38 7.53 20,905 Sprint 7.94 5.05 17,729 T-Mobile 18.07 13.58 28,008 Total 14.13 9.07 126,096 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. Data are for January 2015 – May 2015. Table VI.C.4 FCC Speed Test - Estimated LTE Download Speeds by Service Provider Including Discount Brands, Nationwide382 Service Provider 1H2015 Mean LTE Download Speed (Mbps) Median LTE Download Speed (Mbps) Number of tests Verizon Wireless 16.63 11.66 47,114 AT&T (including Cricket) 10.07 6.84 25,009 Sprint (including Boost, Virgin Mobile) 7.86 4.99 18,280 T-Mobile (including Metro PCS) 17.82 13.27 30,114 Total 14.14 9.07 126,096 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. Data are for January 2015 – May 2015. 380 We intend to release in the near future data with respect to the speeds of the four nationwide service providers’ major discount brands. 381 Estimated download speed data excludes any discount brands offered by a service provider. 382 We recognize that some discount brands may have capped LTE download speeds, or employ different network management practices. See, e.g., CricketWireless website, “Mobile Broadband Information,” available at https://www.cricketwireless.com/legal-info/mobile-broadband-information.html; Virgin Mobile website, available at http://coverage.sprint.com/coverageDescVMU.html. Federal Communications Commission DA 15-1487 85 3. RootMetrics 131. RootMetrics runs a test program that measures mobile data, call, and text performance in all 50 states across the United States.383 According to RootMetrics, tests are performed during all hours of the day, every day of the week, and nearly every week of the year, but the testing schedule is weighted more heavily toward typical consumer usage hours. Performance is measured indoors and outdoors at the same randomly chosen locations, and drive testing takes place during travel between locations. The RootMetrics Speed Index takes into account speed measurements of both data and texts.384 Results are reported at the national, state, and metro levels, and are then combined and converted into scores using a proprietary algorithm. RootScores are meant to reflect a consumer’s experience of network performance and are scaled from 0 – 100,385 with the lower limit representing network performance that would result in a poor consumer experience and the upper limit reflecting extraordinary performance.386 In addition, Root Metrics has provided the Commission with underlying speed numbers that allow us to calculate nationwide LTE download speeds, as presented in Table VI.C.5 below.387 Table VI.C.5 Root Metrics Speed Test – Estimated LTE Download Speeds by Service Provider, Nationwide388 Service Provider 1H2015 Mean Down load Speed (Mbps) Median Download Speed (Mbps) Number of tests Verizon Wireless 15.87 15.68 266,147 AT&T 9.77 9.93 266,120 Sprint 4.80 5.06 265,968 T-Mobile 9.65 9.49 265,789 Total 10.02 10.04 1,064,024 Source: RootMetrics Data, 2015, © Rootmetrics. All rights reserved. In 1H15, there were 6,091,631 total tests, including 7,323 indoor locations. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. 383 Tests are conducted in the 125 most populous metropolitan markets and within the 50 busiest U.S. airports. Each location is tested twice a year, once in the first half of the year and once in the second half of the year. Tests are conducted on the latest Android smartphone available from each service provider. All tests, which are conducted solely on the networks of the four nationwide service providers, are performed identically across all operators’ devices. See RootMetrics website, “Testing Methodology,” available at http://www.rootmetrics.com/us/methodology. In addition to the performance scores at each location, an Online Coverage Map is available (http://webcoveragemap.rootmetrics.com/us). This map incorporates the sample data described above, along with crowdsourced data that are available through consumer use of the free CoverageMap app, available on Android and iOS. 384 See RootMetrics website, “Testing Methodology,” available at http://www.rootmetrics.com/us/methodology. 385 Prior to January 2014, Data RootScores in Metro and Airport RootScore Reports could exceed 100 if performance was extraordinary. See “Testing Methodology,” available at http://www.rootmetrics.com/us/methodology. 386 See RootMetrics website, “Testing Standard,” available at http://www.rootmetrics.com/us/standards. 387 We present additional results in Appendix Tables VI.C.xxi – VI.C.xxvi. 388 Rootmetrics speed data, © Rootmetrics. All rights reserved. Published with permission of Rootmetrics. In addition to the current table, this is applicable to Table VI.C.7 and Appendix Tables VI.C.xxiii-VI.C.xxvi. Federal Communications Commission DA 15-1487 86 4. CalSPEED 132. CalSPEED is an open source, non-proprietary, network performance measurement tool and methodology created for the CPUC with the assistance of a grant from the National Telecommunications and Information Administration (“NTIA”).389 The CalSPEED data presented in this Report are the result of a structured sampling program of nearly 2,000 locations scattered throughout California.390 In our presentation of CalSPEED data, we have dropped any observation that was not in the provider’s coverage area, or any observation that was terminated by the tester. Any other errors are counted as zero throughput.391 Median LTE download speed measurements for the state of California, estimated using CalSPEED data collected during the first half of 2015, are presented in Chart VI.C.6 below. 392 A comparison of median California LTE download speeds estimated using Ookla, FCC, RootMetrics, and CalSPEED data are presented in Table VI.C.7 below. Table VI.C.6 CalSPEED - Estimated LTE Download Speeds, by Service Provider, California Only Service Provider Spring 2015 Mean LTE Download Speed (Mbps) Median LTE Download Speed (Mbps) Number of Tests Verizon Wireless 11.93 11.75 1,387 AT&T 9.06 7.56 1,298 Sprint 5.30 3.36 746 T-Mobile 9.84 9.57 688 Total 9.48 8.34 4,119 Source: The estimated speeds are based on the CalSPEED data. The top 1% of speed values were dropped, by service provider and time period. Spring 2015 tests were taken between the dates of May 1, 2015 through June 15, 2015. 389 We present additional CalSPEED results in Appendix VI., Tables VI.C.xxvii – VI.C.xxx, and also in Charts VI.C.i – VI.C.iii. See also Seventeenth Report, 29 FCC Rcd at 15410 ¶ 201, 15469-71, Appendix VI.C.iv. 390 See CPUC website, “Mobile Broadband Testing,” available at http://www.cpuc.ca.gov/PUC/Telco/bb_drivetest.htm. 391 Upload and download speeds estimated using all technologies are presented in Appendix VI. The rankings are similar to those for LTE speeds. Data can be downloaded from the CPUC website, available at http://www.cpuc.ca.gov/PUC/Telco/bb_drivetest.htm (“Mobile Broadband Testing”). 392 These sites are visited every six months and tests are run on both the latest Android phones and a USB network device on a Windows based netbook, for each of the four nationwide service providers. CalSPEED has now had seven rounds of sampling in California. Data can be downloaded from the CPUC website, available at http://www.cpuc.ca.gov/PUC/Telco/bb_drivetest.htm (“Mobile Broadband Testing”). Federal Communications Commission DA 15-1487 87 Table VI.C.7 LTE Download Speeds by Speed Test and Service Provider, California Only Service Provider Ookla: 1H2015 FCC: 1H2015393 Flagship Brands Only FCC: 1H2015394 Includes Discount Brands Root Metrics: 1H2015 CalSPEED: Spring 2015 Median LTE Down- load speed (Mbps) # of tests (’000s) Median LTE Down- load speed (Mbps) #of tests Median LTE Down- load speed (Mbps) # of tests Median LTE Down- load speed (Mbps) # of tests Median LTE Down- load speed (Mbps) #of Tests Verizon Wireless 15.27 442 10.52 6,740 10.52 6,740 21.44 37,331 11.75 1,387 AT&T 9.57 579 6.51 2,894 6.09 3,238 12.56 37,288 7.56 1,298 Sprint 6.78 295 4.53 1,813 4.46 1,853 6.56 37,347 3.36 746 T-Mobile 12.86 1,101 12.92 5,573 11.8 6,324 13.15 37,210 9.57 688 Total 9.36 2,448 8.64 18,465 8.64 18,466 13.43 149,176 8.34 4,119 * FCC Speed Estimates: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. The FCC data column that includes the discount brands for the four nationwide service providers include the following: T- Mobile includes MetroPCS, AT&T includes Cricket Wireless, and Sprint includes Virgin and Boost. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. ** Ookla Speed Data: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. ***Root Metrics Data: © Rootmetrics. All rights reserved. In 1H15, there were 844,004 total tests, including 993 indoor locations. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. **** CalSPEED: The top 1% of speed values were dropped, by service provider and time period. Spring 2015 tests were taken between the dates of May 1, 2015 to June 15, 2015. 5. Latency 133. Latency refers to several types of delays typically incurred during network data processing, and is typically measured in milliseconds (“ms”). One common measure is round-trip latency, which measures the amount of time it takes a data packet to travel from a source to a destination and back.395 Latency is often affected 393 Estimated download speed data excludes any discount brands offered by a service provider. 394 Consistent with the methodology used in previous reports, estimated download speed data includes any discount brands offered by a service provider. We recognize that some discount brands may have capped LTE download speeds, or employ different network management practices. See, e.g., CricketWireless website, “Mobile Broadband Information,” available at https://www.cricketwireless.com/legal-info/mobile-broadband-information.html; Virgin Mobile Website, available at http://coverage.sprint.com/coverageDescVMU.html 395 More precisely, it is measured as the sum of time from the start of packet transmission by a source to the start of packet reception by a destination plus the time that it takes for the packet to travel back from the receiving destination to the source. This excludes the amount of time that a destination system spends processing the packet. Federal Communications Commission DA 15-1487 88 by factors such as the specifics of the cellular network architecture or processing delays that may occur when the packets need to pass through proxy servers.396 134. FCC Data. Similar to Ookla, the FCC test selects the server with the lowest latency, chosen from an initial set of possible test servers, which may bias subsequent tests towards higher performance. It is possible that consecutive tests in the same place, on the same service provider, and at about the same time may test to different servers. If a packet is not received back within three seconds of sending, it is treated as lost.397 Based on FCC Speed Test App data for 2H2014 and 1H2015, Verizon Wireless had the lowest reported mean latency, followed closely by AT&T and T-Mobile. Sprint had the highest mean latency during the reporting period. 135. CalSPEED Data. CalSPEED tests the complete network path, from the client device, through the local access network, through the Internet backbone, to two ultimate server destinations, one in Northern California and the other in Northern Virginia. Based on the CalSPEED data, latency continues to improve. Verizon Wireless and Sprint have seen dramatic improvements in their latency over the last several reporting periods, and are closing in on AT&T for the lowest latency measurements. While T-Mobile has the highest mean latency, 398 they have also demonstrated dramatic improvements during the last several test periods. D. Differentiation in Mobile Wireless Devices and Advertising/Marketing 1. Differentiation in Mobile Wireless Devices 136. In addition to competing on price and network quality, service providers have also competed by offering consumers a variety of different mobile wireless devices with innovative features. Today, service providers offer a range of data-centric smartphones399 and tablets made by different manufacturers with different operating systems, and increasingly offer wearable devices, such as smartwatches.400 In many cases, customers may be able to find the same devices from several different service providers, but the promotion and marketing of specific devices may vary across providers. The first significant differentiation in mobile device offerings occurred with the introduction of Apple’s iPhone in June 2007 pursuant to an exclusive arrangement with AT&T. Following that introduction, many handset manufacturers introduced competing products with similar features such as touch screens, mobile web browsing capabilities, and current-generation operating systems. 137. Today, smartphone operating systems such as the Android and the Apple iOS are available from multiple service providers, permitting consumers to pair their preferred operating systems with various service providers. While certain specific smartphones may be introduced by a single provider, the wide variety of similar models has made it more difficult for service providers to differentiate their services based upon differences in device offerings alone. We also note that all four nationwide service providers, as well as many regional and 396 See FCC website, International Broadband Data Report, DA 12-1334, at 14, available at http://www.fcc.gov/document/international-broadband-data-report. 397 See FCC website, “Measuring Mobile Broadband Methodology,” available at http://www.fcc.gov/measuring-broadband- america/mobile/technical-summary. 398 See “CALSpeed – California’s Mobile Broadband Assessment” (Spring 2015). 399 While there is no industry standard definition of a smartphone, for purposes of this Report we continue to consider the distinguishing features of a smartphone to be: an HTML browser that allows easy access to the full, open Internet; an operating system that provides a standardized interface and platform for application developers; and a larger screen size than a traditional, voice-centric handset. Many smartphones also have touch screens and/or a QWERTY keypad, and run an operating system that offers a standard platform for application developers to create and sell device software through an application store. See Sixteenth Report, 28 FCC Rcd at 3844 ¶ 220. By contrast, the basic handset category includes voice- centric handsets that do not allow or are not designed for easy web browsing. In addition to smartphones and basic handsets, a third category of devices consists of data-centric devices that have no inherent voice capability, such as USB wireless modem laptop cards, mobile Wi-Fi devices, e-readers, and laptops and netbooks with embedded mobile wireless modems. 400 See “The Trick for Wearables’ Success is all in the Wristware,” Mar. 30, 2015, available at http://www.cnet.com/news/wearables-set-to-soar-on-the-back-of-wristwear-new-data-shows/. Federal Communications Commission DA 15-1487 89 smaller service providers, currently offer the current iPhone models. Although all service providers continue to offer the highest end devices only on postpaid plans, providers are increasingly targeting prepaid customers by offering more devices with prepaid plans than before. Beginning in 2015, Walmart also began offering Sprint’s Virgin prepaid shared data plans with purchase of the HTC Desire 510, the LG Tribute, the LG Volt, and the Samsung Galaxy Core Prime smartphones.401 This offering was the first prepaid plan to allow for shared data.402 138. In addition to offering more devices on prepaid plans, service providers have been promoting cheaper tablets,403 in part to generate growth in data traffic and generate a market for future video offerings.404 Internet device net adds (including tablets) were 1.97 million in the second quarter of 2015, down from 2.27 million in the second quarter of 2014, and this marks the first decline in three years.405 AT&T increased its net internet device adds from 449,000 in the first quarter of 2014 to 711,000 in the first quarter of 2015, and saw a decline in the second quarter of 2015, to 600,000.406 Verizon Wireless increased its net internet device adds from 621,000 in the first quarter of 2014 to 667,000 in the first quarter of 2015, and to 803,000 in the second quarer of 2015.407 Sprint saw a decrease in its net internet device adds from 414,000 in the first quarter of 2014 to 379,000 in the first quarter of 2015 to 322,000 in the second quarter of 2015.408 T-Mobile had the fewest internet device net adds (134,000 in the first quarter of 2015, and 248,000 in the second quarter of 2015) as it entered the tablet market later than the other service providers.409 139. Although the use of data-only devices with mobile network connectivity has grown in recent years, “phablets,” a class of mobile device combining the form and technical capabilities of smartphones and tablets, are replacing the use of tablets in some cases.410 Phablet sales have been substantial and it has been predicted that phablets will account for more than 20 percent of mobile devices shipped by 2019.411 In part, due to the popularity of phablets, Apple launched the iPhone 6 Plus last year to reclaim customers lost to Android 401 See “Sprint’s Virgin Prepaid Service to Offer Shared Data Plans at Walmart,” Jan. 16, 2015, available at http://www.cnet.com/news/sprints-virgin-prepaid-service-to-offer-shared-data-plans-at-walmart/. 402 See id. 403 According to Cisco VNI Mobile, in 2014, a 4G smartphone, on average, used approximately 2 GB per month, as compared to 2.9 GB for a 4G tablet. Cisco predicts that by 2019, this is expected to grow to approximately 4 GB per month for a 4G smartphone, as compared to approximately 12 GB per month for a 4G tablet. See Cisco VNI Mobile, “Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update 2014-2019 White Paper,” at 19, available at http://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/white_paper_c11- 520862.html. 404 See UBS, US Wireless 411: Version 56, May 14, 2015, at p. 11. 405 See UBS, US Wireless 411: Version 57, Aug. 17, 2015, at p. 20. 406 See UBS, US Wireless 411: Version 56, May 14, 2015, at p. 11; US Wireless 411: Version 57, Aug. 17, 2015, at p. 20. 407 See id. 408 See id. 409 See id.; see also “T-Mobile Revolutionizes how Customers Buy and Use Tablets with Free Data for Life,” available at https://newsroom.t-mobile.com/news/t-mobile-revolutionizes-how-customers-buy-and-use-tablets-with-free-data-for- life.htm. 410 The phablet has been defined as a handset with a screen size of 5.5 to 6.9 inches that can be held while making phone calls, but not necessarily for an extended period of time. See “Phablets to Flood Smartphone Market in Coming Years,” Jan. 28, 2015, available at http://www.cnet.com/news/phablets-to-flood-smartphone-market-in-coming-years-report/. 411 See id. Federal Communications Commission DA 15-1487 90 vendors.412 In the first quarter of 2015, 44 percent of all U.S. phablet sales were the iPhone 6 Plus.413 In addition to offering a variety of smartphones, traditional handsets, and phablets, mobile wireless service providers also sell or provide connectivity for other data-only devices such as tablets, e-readers, wireless data cards, mobile Wi-Fi hotspots,414 netbook computers with embedded modems, and wearable gadgets such as smartwatches. Smartwatches such as Android watches have been available for several years, but they have become more popular since Apple introduced its Apple Watch in April 2015. Mobile wireless service providers also offer wireless data cards and mobile Wi-Fi hotspots to consumers seeking mobile Internet connections for laptop computers and other Wi-Fi enabled devices and may offer products for the interconnected home, such as security systems. 140. As service providers have become more able recently to compete with one another by offering many devices with similar characteristics from multiple equipment manufacturers, providers have sought to compete more along other dimensions, such as network speed and coverage. That being said, service providers may emphasize and promote different smartphone brands, although other brands may be available in their stores. According to a JD Power report on consumer satisfaction with smartphones, the primary reasons for purchasing a smartphone device differ by service provider. For example, Verizon Wireless customers are more likely to buy their device based on the smartphone’s features (approximately 34 percent), while T-Mobile customers are more likely to buy based on price (approximately 30 percent).415 2. Advertising and Marketing 141. Mobile wireless service providers also compete for customers through advertising and marketing, including by establishing retail and distribution networks. Several service providers stated that the goal of their advertising and marketing efforts is to increase and maintain brand awareness and to support distribution.416 Service providers may also engage in advertising and marketing either to inform consumers about available products or services or to try to increase sales by influencing consumer preferences.417 Service providers may advertise through traditional media, internet and mobile applications, social media, sponsorships and co-branding, and at events.418 During the time period covered by this Report, the four nationwide service providers all increased their advertising expenditures, some quite substantially,419 and telecom companies overall spent more on advertising than all but two industries in the first quarter of 2015.420 According to Neilsen, wireless telephone 412 See id. 413 See “iPhone 6 Plus Leads Phablet Sales in U.S.,” May 6, 2015, available at, http://www.cnet.com/news/iphone-6-plus- leads-phablet-sales-in-us/. 414 Mobile Wi-Fi, or “Mi-Fi,” devices are credit card-sized, mobile Wi-Fi routers with mobile broadband wide-area connections that allow a certain number of Wi-Fi-enabled devices in short range to connect to the Internet via a Wi-Fi connection. Many smartphones are now sold with built-in Wi-Fi hotspot capabilities, allowing them to serve as mobile Wi-Fi hotspots for an additional charge. 415 See “Smartphone Device Launches Propel High Satisfaction with Apple, HTC, and Samsung,” Apr. 23, 2015, available at http://www.jdpower.com/press-releases/2015-us-wireless-smartphone-satisfaction-study%E2%80%94volume-1#. 416 See, e.g., 2014 SEC Form 10-K for Sprint at 3, 2014 SEC Form 10-K for US Cellular at 2-3, 2014 SEC Form 10-K for Verizon Wireless at 6. 417 See Seventeenth Report, 29 FCC Rcd at 15140 ¶ 212. 418 See, e.g., 2014 SEC Form 10-K for Sprint, 2014 SEC Form 10-K for US Cellular, and 2014 SEC Form 10-K for Verizon Wireless. 419 CTIA Comments at 47-48 (citing Kantar Media Reports). AT&T reported advertising costs in 2014 as $3.3 billion; Sprint reported sales and marketing expense was $5.3 billion for the year ended Mar. 31, 2015 representing an increase of $2.7 billion, or 102%; T-Mobile reported advertising expense included in selling, general and administrative expenses as $1.4 billion for 2014; and Verizon Wireless reported advertising expenses at $2.5 billion in 2014. 420 CTIA Comments at 47 (citing Kantar Media Reports). Federal Communications Commission DA 15-1487 91 services was the sixth largest category of U.S. advertising spending ($2.2 billion from Jan. 1, 2014 to Sept. 30, 2015). In Kantar Media’s Q3 2014 rankings of advertising spending, AT&T and Verizon Communications were the second and eighth largest U.S. advertisers, respectively,421 although T-Mobile outspent the other wireless providers (including AT&T and Verizon Wireless) in July 2015.422 142. Some service providers’ marketing campaigns continued to focus on the quality and size of their mobile broadband networks. Many service providers sought to highlight their network speed, coverage, reliability, and the data capabilities of devices available on these networks.423 Some service providers promoted the advantages of their particular service plans relative to those of rivals, including phone upgrade options and loyalty rewards programs. For example, in 2015, AT&T partnered with American Express and some other major brands to offer a loyalty program for customers called Plenti, which enabled consumers to shop at any of the stores in the program to earn points toward rewards.424 In 2015, Verizon Wireless used as an advertising slogan, the “never settle” tag line, implying that customers who choose other providers are settling for inferior network quality.425 In response, in May 2015, T-Mobile used the tag line #NeverSettleForVerizon and offered Verizon Wireless customers the chance to port their number to a new T-Mobile smartphone on a trial basis.426 In June 2015, T-Mobile announced its “Jump on Demand” marketing slogan which advertised that customers would be allowed to upgrade to new phones up to three times per year.427 In the summer of 2015, Verizon Wireless aired an ad that featured a flock of geese who are “flying back home to Verizon” after having “suffered” after leaving Verizon Wireless because of other wireless service providers’ “unreliable cell phone networks and videos that kept buffering.”428 T-Mobile aired its own ad featuring geese, informing the audience that it has expanded LTE coverage, and that it now has improved coverage and in-building penetration.429 VII. CONSUMERS AND TRENDS IN THE MOBILE WIRELESS ECOSYSTEM 143. In today’s connected world, consumers are faced with a wide variety of choices in mobile service plans, devices and applications.430 In this section, we first report on trends in handsets and smartphone 421 See “Kantar Media Reports U.S. Advertising Expenditures Increased 0.3 Percent in Q3 2014, Helped by the World Cup and Political Spend,” Dec. 22, 2014, available at http://kantarmedia.us/kantar-media-reports-us-adex-increase-q3-2014. 422 See “The Top 5 Wireless Adds: T-Mobile Outspent AT&T, Verizon, and Sprint in TV Advertisements in July,” Aug. 12, 2015, available at http://www.fiercewireless.com/special-reports/top-5-wireless-ads-t-mobile-outspent-att-verizon-and-sprint- tv-advertisemen?confirmation=123. 423 See Verizon Wireless Comments at 13. 424 See “AT&T and Friends Form Rewards Program,” Mar. 18, 2015, available at http://www.lightreading.com/spit-(service- provider-it)/customer-experience-management-(cem)/atandt-and-friends-form-rewards-program/d/d-id/714502. 425 See “T-Mobile Targets Verizon Customers with Free Two Week Trial,” May 5, 2015, available at http://www.fiercewireless.com/story/t-mobile-targets-verizon-customers-free-two-week-trial/2015-05-05. 426 See id. 427 See “T-Mobile Will Let You Upgrade Your Smartphone Any Time You Want,” June 25, 2015, available at http://www.cnet.com/news/t-mobile-will-let-you-upgrade-your-smartphone-anytime-you-want/. 428 See “Verizon TV Spot, ‘Magnificent Geese,’” available at http://www.ispot.tv/ad/7PpP/verizon-magnificent-geese. 429 See “T-Mobile jabs at Verizon in new ad that highlights its 700 MHz LTE coverage expansion,” available at http://www.fiercewireless.com/story/t-mobile-jabs-verizon-new-ad-highlights-its-700-mhz-lte-coverage-expansion/2015-09- 22. 430 Consumers choose a service provider or switch between providers for varying reasons, including price, availability of family plans, network quality, free/unlimited in-network calling, billing/payment options/credit, reputation/recommendation, previous experience with the service provider, customer service, mobile data services, specific phone offerings, bundling mobile phone services with other services or other unspecified reasons. In the past, contract length, handset exclusivity, and lack of interoperability were some factors that were highlighted as barriers to switching. Consumers may incur some Federal Communications Commission DA 15-1487 92 penetration, considering the handset operating system, and application (“apps”) dimensions of the mobile wireless ecosystem. Next, we consider trends in consumers’ use of mobile wireless services. The handsets offered by, or compatible with a particular service provider, as well as the available apps, will affect consumers’ decisions since each greatly affects the user experience, as discussed below. This section then discusses consumers’ access to information, and lastly, we report on trends in consumers’ use of mobile versus non-mobile services. A. Handsets and Downstream Mobile Applications 144. Smartphone Penetration. Smartphone use has continued to increase over the last three years, as Chart VII.A.1 shows, although smartphone use flattened out in 2015. According to ComScore, approximately 77 percent of all mobile subscribers had a smartphone in the third quarter of 2015, compared to approximately 51 percent in the third quarter of 2012.431 The smartphone penetration rate among new phone purchases over the same period stood at approximately 88 percent in the third quarter of 2015 (again flattening out in 2015), up from approximately 67 percent in the third quarter of 2012. Among the top four mobile wireless service providers, the penetration rates for postpaid smartphone subscribers are approximately 94 percent, 89 percent, 83 percent, and 80 percent for T-Mobile, Sprint, AT&T, and Verizon Wireless, respectively.432 Note: ComScore, MobiLens Audience Profile, 3-month averages from Q3 2012 to Q3 2015. 145. Share of Smartphones by Operating System. The operating system (“OS”) of a smartphone is a major determining factor of the smartphone’s ability to support mobile applications and Internet-based services. As seen in Chart VII.A.2, Apple’s iOS and Google’s Android continued to lead the market for mobile operating systems. In the third quarter of 2015, Android’s operating system accounted for approximately 52 percent of the smartphone OS market, while Apple’s iOS accounted for approximately 44 percent. While Android’s market share has been fairly constant since the third quarter of 2012, Apple’s market share has grown from approximately 34 percent to approximately 44 percent over the same time period. RIM (Research in switching costs, even in today’s mobile wireless marketplace, including search costs, handset purchases, and implicit costs such as brand loyalty. See Seventeenth Report, FCC Rcd at 15347 ¶ 69. 431 See ComScore, MobiLens Trend, 3-month averages from August 2012 to September 2015. 432 See UBS, US Wireless 411, Version 57, Aug. 2015, at p. 26, Figure 37. 5 0 .0 % 5 1 .0 % 5 2 .0 % 5 3 .0 % 5 4 .0 % 5 5 .0 % 5 6 .0 % 5 8 .0 % 5 8 .0 % 5 9 .0 % 5 9 .0 % 6 0 .0 % 6 1 .0 % 6 2 .0 % 6 3 .0 % 6 4 .0 % 6 5 .0 % 6 7 .0 % 6 8 .0 % 6 9 .0 % 7 0 .0 % 7 0 .0 % 7 2 .0 % 7 2 .0 % 7 2 .0 % 7 2 .0 % 7 3 .0 % 7 4 .0 % 7 5 .0 % 7 6 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 7 7 .0 % 6 4 .0 % 6 7 .0 % 6 7 .0 % 7 0 .0 % 7 2 .0 % 7 3 .0 % 7 3 .0 % 7 2 .0 % 7 2 .0 % 7 4 .0 % 7 5 .0 % 7 6 .0 % 7 5 .0 % 7 6 .0 % 7 7 .0 % 8 0 .0 % 8 3 .0 % 8 4 .0 % 8 3 .0 % 8 1 .0 % 8 2 .0 % 8 3 .0 % 8 5 .0 % 8 4 .0 % 8 4 .0 % 8 5 .0 % 8 6 .0 % 8 8 .0 % 8 7 .0 % 8 8 .0 % 8 7 .0 % 8 9 .0 % 8 8 .0 % 8 8 .0 % 8 7 .0 % 8 8 .0 % 8 7 .0 % 8 8 .0 % 0% 20% 40% 60% 80% 100% A u g -1 2 S ep -1 2 O ct-1 2 N o v -1 2 D ec-1 2 J a n -1 3 F eb -1 3 M a r-1 3 A p r-1 3 M a y -1 3 J u n -1 3 J u l-1 3 A u g -1 3 S ep -1 3 O ct-1 3 N o v -1 3 D ec-1 3 J a n -1 4 F eb -1 4 M a r-1 4 A p r-1 4 M a y -1 4 J u n -1 4 J u l-1 4 A u g -1 4 S ep -1 4 O ct-1 4 N o v -1 4 D ec-1 4 J a n -1 5 F eb -1 5 M a r-1 5 A p r-1 5 M a y -1 5 J u n -1 5 J u l-1 5 A u g -1 5 S ep -1 5 Chart VII.A.1 Smartphone Penetration Rates in the US (3-Month Average) August 2012 - September 2015 Of all Mobile Phones Of recent Purchases (Last Month) Federal Communications Commission DA 15-1487 93 Motion)/Blackberry (approximately one percent), Microsoft (approximately three percent), and other firms (less than one percent) comprised the remainder of the market in the third quarter of 2015. Note: Based on ComScore MobiLens 3-month survey data averages. 146. Smartphones and Consumer Satisfaction. According to an April 2015 JD Power study, consumer satisfaction with smartphone brands differs by wireless service provider.433 In addition, the primary reasons that customers cite for purchasing a smartphone differ by wireless service provider. For example, Verizon Wireless customers are more likely to cite phone features as the primary factor in their purchasing decision, while T- Mobile customers value price most highly. Apple and Samsung smartphones rank highest in satisfaction among customers of T-Mobile, Verizon Wireless, and AT&T, while HTC, Apple, and Samsung rank highest among customers of Sprint.434 Overall satisfaction with smart phone devices is highest among AT&T customers, according to the survey, followed by Sprint, Verizon Wireless, and T-Mobile customers.435 147. Downstream Mobile Applications. Smartphone users generally interact with their devices through specific apps and the increasing use of smartphones has spawned a large and growing mobile app ecosystem. Two app stores dominate the United States market – Google Play and Apple App Store. Google Play had approximately 1.5 million apps as of the first week of May 2015,436 and Apple App Store offered more than 1.4 million apps at the end of 2014.437 Other application stores include the Windows Phone Store, the Amazon 433 See “Smartphone Device Launches Propel High Satisfaction with Apple, HTC, and Samsung,” Apr. 23, 2015, available at http://www.jdpower.com/press-releases/2015-us-wireless-smartphone-satisfaction-study%E2%80%94volume-1#. 434 See id. 435 See id. 436 See “Android Statistics,” available at http://www.appbrain.com/stats/number-of-android-apps. 437 See “App Store Rings in 2015 with New Records,” Apple press release, Jan. 8, 2015. 18.9% 8.4% 3.8% 2.3% 1.2% 44.8% 52.2% 51.8% 52.1% 52.3% 27.4% 34.3% 40.6% 41.7% 43.6% 3.3% 1.2% 0.5% 0.2% 0.1% 5.6% 3.6% 3.3% 3.6% 2.9% 0% 10% 20% 30% 40% 50% 60% Sep-11 Sep-12 Sep-13 Sep-14 Sep-15 Chart VII.A.2 Market Share by Smartphone in Use, By OS (2010- 2015) RIM/BlackBerry Google/Android Apple All Others Microsoft Federal Communications Commission DA 15-1487 94 App Store, and BlackBerry World.438 Mobile applications generate revenue through contracts for application developers, e-commerce sales, in-application advertising, and application store sales.439 Apple App Store generated approximately $10 billion in revenue to developers in 2014, while Google Play has paid out $7 billion to app developers over the same period.440 It is anticipated that app revenue from the Google Play marketplace and Apple App Store is likely to double by 2018.441 Chart VII.A.3 shows the Top 15 types of mobile apps and websites accessed by smartphone users as of September 2015. Note: Based on ComScore MobiLens 3-month survey data averages. B. Usage 148. According to CTIA, reported annual minutes of use (“MOUs”) in 2014 reached over 2.45 trillion, a decrease of 6.2 percent, while average billable minutes of use, as shown in Chart VII.B.1, decreased by approximately 10 percent compared to year-end 2013.442 This may be due in part to changes in how service providers report, service provider participation, and possible reported volume increases/decreases in usage. In addition, as service providers have continued to introduce new all-inclusive calling and data plans, as well as modifications on legacy plans, the reporting of specific breakout data has become more complicated and difficult to report.443 Further, the Internet of Things such as connected cars, telematics and M2M are likely to be harder to 438 For example, Microsoft has more than 585K apps in the Windows and Windows Phone store combined (http://news.microsoft.com/bythenumbers/index.html). As of January 2015, more than 293K mobile apps were available in the Amazon Appstore (http://www.statista.com/statistics/307330/number-of-available-apps-in-the-amazon-appstore/). 439 See Vision Mobile, “Business and Productivity Apps,” March 2014. 440 See Ina Fried, “Sales from Apple and Google App Stores Seen Doubling by 2018, In-App Advertising to Triple,” Mar. 30, 2015, available at http://recode.net/2015/03/30/sales-from-apple-and-google-app-stores-seen-doubling-by-2018-in-app- advertising-to-triple/. 441 See id. 442 See CTIA Wireless Industry Indices Year-End 2014, at p. 125, Table 44: Calendar Year Annualized MOU Data. 443 See id., at pp.131-32 (describing understatement of MOUs). 88.9% 83.0% 75.5% 67.4% 66.3% 60.0% 56.3% 55.0% 54.9% 52.1% 50.9% 50.2% 46.3% 45.3% 42.7% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% P er ce n t o f U se rs Top Categories of Mobile Apps and Websites Chart VII.A.3 Top 15 Types of Mobile Apps and Websites Accessed by Smartphone Users, Sept. 2015 Federal Communications Commission DA 15-1487 95 report consistently across all service providers as many of the providers may not be able to segregate traditional voice usage since it may be included within their total data usage metrics.444 Source: CTIA Wireless Industry Indices Year-End 2014. 149. Although MOUs for voice may have decreased, data use has grown significantly. CTIA reported that SMS and text messaging traffic amounted to over 169.3 billion messages for the December 2014 period, an increase of 10 percent from the prior year, and monthly data usage per smartphone subscriber in 2014 averaged 1.4 GB per month, increasing 18 percent over 2013.445 Total wireless data traffic reported by the service providers to CTIA amounted to 4.06 trillion MB for 2014 up 25.7 percent from 3.23 trillion MB in 2013.446 Chart VII.B.2 provides average data usage per subscriber for 2010 to 2014 comparing the amount of data usage between data-capable devices and smartphones. Source: CTIA Wireless Industry Indices Year-End 2014, Chart 32 Indices. Latest available data. 444 See section II.B.1 supra. 445 See CTIA Wireless Industry Indices Year-End 2014, at p. 136. 446 See id., at p. 11. 758 717 672 675 746 671 726 683 650 633 690 610 0 100 200 300 400 500 600 700 800 2009 2010 2011 2012 2013 2014 A v er ag e M O U s p er S u b sc ri b er p er M o n th Year Chart VII.B.1 Average MOU per Subscriber per Month, 2009-2014 Excludes most data-only devices Includes non-voice capable units 122 257 407 849 1008 269 510 768 1152 1361 0 500 1000 1500 2010 2011 2012 2013 2014 A v er ag e M o n th ly M B p er A v er ag e D at a C ap ab le U n it Chart VII.B.2 Mobile Data Usage per Subscriber, 2010 - 2014 Data Capable Unit Smartphone Federal Communications Commission DA 15-1487 96 150. Further, Cisco’s Visual Networking Index reported that in 2014 the average usage per connection in North America was approximately 1.5 GB.447 Cisco estimated that in North America, mobile data grew 63 percent for 2014.448 Ericsson, in its latest June 2015 North American Mobility report indicated that data traffic per active smartphone user equaled approximately 2.4 GB per month,449 while Chetan Sharma Consulting indicated that the average usage was approximately 2.5 GB per month.450 According to GSMA, in 2014, globally LTE users used twice as much data as non-LTE users, which translated to about 1.5 GB of data per month on average.451 This trend in increasing data use is due to multiple factors, including the increased adoption of smartphones and tablets, growth in streaming video, and the development of faster networks. According to the Pew Research Internet Project, 97 percent of cellphone users use their cellphone to send or receive text messages; 89 percent access the Internet; 88 percent send or receive email; 67 percent get directions, recommendations, or other location-based information; 64 percent of younger smartphone users listen to music; and 91 percent of users between 18 and 29 years of age used social networking compared to 55 percent of those age 50 and older.452 C. Consumer Access to Information 151. Through the “Consumer Code for Wireless Service,” CTIA and the service providers voluntarily commit to providing consumers with information to assist them in the selection of a mobile wireless service provider.453 Signatories to CTIA’s Consumer Code commit to disclose rates, additional taxes, fees, surcharges, and terms of service; provide coverage maps; and make customer service readily accessible. Since its creation, the code has been updated to require providers to ensure disclosure of data allowances offered in a service plan, whether there are any prohibitions on data service usage, and whether there are network management practices that will have a material impact on the customer’s wireless data experience. 152. In December 2013, CTIA added a section enhancing transparency and disclosure of mobile wireless providers’ device unlocking policies.454 The ability to unlock a handset in order to activate it on another service providers network enables consumers to exercise greater choice in choosing or switching providers and lowers switching costs. These requirements include notifying customers when their postpaid device is eligible for unlocking if the device is not automatically unlocked. In addition, participating service providers are required to 447 Cisco’s Visual Networking Index, reported that in 2014 the average usage per connection in North America was 1,477 megabytes. See Cisco Visual Networking Index: Forecast and Methodology, 2014-2019 (Feb. 3, 2015): average mobile connections in 2019 will generate 7,648 megabytes of mobile data traffic per month. 448 See id. Global mobile data traffic is predicted to increase nearly tenfold between 2014 and 2019. Mobile data traffic will grow at a compound annual growth rate (“CAGR”) of 57% from 2014 to 2019, reaching 24.3 exabytes per month by 2019. 449 See Ovum: “Smartphone & tablet usage trends & insights: 4G LTE and Wi-Fi powering data consumption” (2015). Ericsson – a network infrastructure service provider – projects mobile data growth of 35% per year between 2014 and 2020. See Ericsson Mobility Appendix (June 2015). 450 See “Industry Research: Technology & Strategy Consulting,” available at http://www.chetansharma.com/research.htm. See also “Report: U.S. Consumers Swallowed 2.5 GB/month of Cellular Data in Q1 on Average,” available at http://www.fiercewireless.com/story/report-us-consumers-swallowed-25-gbmonth-cellular-data-q1-average/2015-05-19. 451 See GSMA Report, “The Mobile Economy 2014,” available at http://www.gsmamobileeconomy.com/GSMA_ME_Report_2014_R2_WEB.pdf. 452 See “U.S. Smartphone Use in 2015” (Apr. 1, 2015), available at http://www.pewinternet.org/files/2015/03/PI_Smartphones_0401151.pdf. 453 See “Consumer Code for Wireless Service,” available at http://www.ctia.org/policy-initiatives/voluntary- guidelines/consumer-code-for-wireless-service. 454 See id. Federal Communications Commission DA 15-1487 97 post on their websites a clear, concise, and easily found policy on mobile device unlocking.455 The member service providers were required to implement all of the unlocking disclosure policies by February 11, 2015, and all of the major service providers have fulfilled this commitment.456 The member service providers agreeing to CTIA’s Consumer Code account for service to 97 percent of U.S. wireless customers.457 All of these customers must be sent alerts regarding data, voice, text, international roaming, and device unlocking eligibility, unless they decide to opt out. In order to further facilitate the adoption of such alerts, the Commission has established a web site where consumers can determine which providers are implementing the voluntary commitments.458 153. Open Internet Rules. The rules on Internet openness adopted by the Commission in February 2015, require both fixed and mobile broadband Internet service providers to “publicly disclose accurate information regarding the network management practices, performance, and commercial terms of its broadband Internet access services sufficient for consumers to make informed choices regarding use of such services.”459 The 2015 Open Internet Order reaffirms and enhances transparency rules that were originally adopted in 2010.460 In 2014, the D.C. Circuit Court rejected a challenge by Verizon to the 2010 transparency rule, although it did invalidate anti-blocking and anti-discrimination rules.461 Subsequently, in 2015, the Commission adopted new rules against blocking, throttling, and paid prioritization for fixed and mobile service providers.462 154. In providing guidance regarding effective disclosure models in the 2010 and 2015 Orders, the Commission indicated that among the types of information that might be included in an effective disclosure are pricing terms such as monthly prices, usage-based fees, and fees for early termination or additional network services.463 The 2015 Order enhanced the transparency rules adopted in 2010 by adopting a requirement that broadband service providers always must disclose promotional rates, all fees and/or surcharges, and all data caps or data allowances.464 In addition, the 2015 Order requires that packet loss as a measure of network performance 455 See FCC website, Official FCC Blog, “Wireless Providers Fulfill Commitment to Let Consumers Unlock Mobile Phones,” available at https://www.fcc.gov/blog/wireless-providers-fulfill-commitment-let-consumers-unlock-mobile-phones. 456 See “Consumer Code for Wireless Service,” available at http://www.ctia.org/policy-initiatives/voluntary- guidelines/consumer-code-for-wireless-service. 457 See id. 458 See FCC website, “Helping Consumers Avoid Bill Shock,” available at https://www.fcc.gov/bill-shock-alerts. 459 Protecting and Promoting the Open Internet, GN Docket No. 14-28, Report and Order on Remand, Declaratory Ruling, and Order, 30 FCC Rcd 5601, 5609 ¶ 23 (2015) (“2015 Open Internet Order”). 460 See id. See also Preserving the Open Internet, GN Docket No. 09-191, WC Docket No. 07-52, Report and Order, 25 FCC Rcd 17905, 17938-39 ¶ 56 (2010) (“2010 Open Internet Order”). 461 See Verizon v. Federal Communications Commission, 740 F.3d 623 (D.C. Cir. 2014). 462 See 2015 Open Internet Order, 30 FCC Rcd at 5607-8 ¶¶ 15-19. 463 See 2015 Open Internet Order, 30 FCC Rcd at 5609 ¶ 23; 2010 Open Internet Order, 25 FCC Rcd at 17938-39 ¶ 56. On July 23, 2014, the Enforcement Bureau of the FCC announced that, “Providers of broadband Internet access services must disclose accurate information about their service offerings and make this information accessible to the public.” (See DA 14- 39, Enforcement Advisory No. 2014-03, July 23, 2014, (https://transition.fcc.gov/eb/Public_Notices/DA-14-1039A1.html)). The Commission charged its Consumer Advisory Committee with making recommendations regarding proposed consumer facing disclosure format, based on input from a broad range of stakeholders, accessible to persons with disabilities, and to consider whether different formats should be used for fixed and mobile service providers. See 2015 Open Internet Order, 30 FCC Rcd at 5680-81 ¶ 180. The recommendations were submitted on October 26, 2015, and are available on the FCC website, “Consumer Advisory Committee Recommendations 2014 thru 2016,” https://www.fcc.gov/consumer-advisory- committee-recommendations-2014-thru-2016. 464 See 2015 Open Internet Order, 30 FCC Rcd at 5609 ¶ 23. Federal Communications Commission DA 15-1487 98 must be disclosed and requires specific notification to consumers that a “network practice” is likely to significantly affect their use of the service.465 D. Intermodal Developments 155. Advances in technologies and functionalities have made mobile broadband services more versatile and useful to consumers. However, while fixed and mobile broadband services may provide overlapping capabilities, mobile and fixed services are not co-extensive in their capabilities; there are unique capabilities for each of fixed and mobile services. It is also sometimes the case that mobile services and fixed services enhance the quality of one another. In fact, residential and business consumers alike often use mobile and fixed services in concert to, for example, off load reliance from cellular networks to Wi-Fi systems that are connected to the internet via a fixed service.466 In addition, the increasingly dynamic nature of residential and business communication requires a mix of fixed and mobile broadband access to provide sufficient functionality for families and businesses whose members often simultaneously rely on data-capacity intensive applications at fixed locations and mobile applications on the go. 156. For voice services, we here provide the latest information from the Centers for Disease Control and Prevention’s (“CDC”) National Health Interview Survey on wireless substitution, which we emphasize only pertains to voice services and therefore no inferences regarding broadband services can, or should, be drawn based upon it. The survey includes information about household telephones and whether anyone in the household has a wireless telephone. Preliminary results from CDC’s January to June 2015 National Health Interview Survey indicate that the number of American homes with only wireless telephones continues to grow. As shown in Chart VII.D.1 and Chart VII.D.2, the percentage of U.S. adults and children living in households with landlines, with or without wireless, has fallen steadily over the past several years.467 The percentage of wireless-only households has continued to increase for both groups, and the percentage of households without phones has increased slightly.468 A significant percentage of homes with both landline and wireless phone access received all or almost all calls on wireless telephones despite also having a landline telephone.469 465 See id. 466 According to Cisco VNI Mobile, “By 2016, more than half of all traffic from mobile-connected devices will be offloaded to the fixed network by means of Wi-Fi devices and femtocells each month.” See Cisco VNI Mobile, “Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update 2014-2019 White Paper,” at 4, available at http://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/white_paper_c11- 520862.html. As of 2014, 45% of total mobile data traffic from all mobile-connected devices was offloaded. See id., at 22. 467 See Appendix Tables VII.C.i and VII.C.ii. 468 See Stephen J. Blumberg and Julian V. Luke, “Wireless Substitution: Early Release of Estimates from the National Health Interview Survey,” January to June 2015, National Center for Health Statistics, Centers for Disease Control (December 2015), available at http://www.cdc.gov/nchs/data/nhis/earlyrelease/wireless201512.pdf. 469 See id. Federal Communications Commission DA 15-1487 99 62.2% 59.4% 58.8% 57.3% 56.1% 54.4% 52.8% 51.5% 47.3% 45.8% 43.9% 24.9% 27.8% 30.2% 32.3% 34.0% 36.5% 38.0% 39.1% 43.1% 44.1% 46.7% 10.9% 10.7% 9.0% 8.3% 7.8% 7.0% 6.9% 7.0% 7.0% 7.1% 6.2% 1.7% 1.8% 1.8% 1.9% 1.9% 1.9% 2.2% 2.2% 2.4% 2.9% 3.1% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Jan–Jun 2010 Jul–Dec 2010 Jan–Jun 2011 Jul–Dec 2011 Jan–Jun 2012 Jul–Dec 2012 Jan–Jun 2013 Jul-Dec 2013 Jan-Jun 2014 Jul-Dec 2014 Jan-Jun 2015 P er ce n t Chart VII.D.1 Percentage of U.S. Adults Living in Households with/without Wireless and Landlines (2010 - June 2015) Landline with Wireless Wireless-only Landline without Wireless Phoneless Landline with Wireless Wireless only Landline without Wireless Phoneless Note: Adults are aged 18 and over, children are under age 18 Source: CDC/NCHS National Health Interview Survey 62.8% 59.8% 56.7% 54.7% 52.7% 49.5% 48.3% 46.4% 47.1% 39.1% 38.3% 29.0% 31.8% 36.4% 38.1% 40.6% 45.0% 45.4% 47.1% 52.1% 54.1% 55.3% 6.4% 6.2% 5.1% 4.8% 4.5% 3.4% 3.6% 3.8% 3.5% 3.3% 3.0% 1.7% 2.0% 1.7% 2.2% 2.2% 1.9% 2.6% 2.5% 2.7% 3.4% 3.2% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Jan – Jun 2010 Jul–Dec 2010 Jan–Jun 2011 Jul–Dec 2011 Jan–Jun 2012 Jul–Dec 2012 Jan–Jun 2013 Jul - Dec 2013 Jan-Jun 2014 Jul-Dec 2014 Jan-Jun 2015 P er ce n t Chart VII.D.2 Percentage of U.S. Children Living in Households with/without Wireless and Landlines, 2010 - June 2015 Landline with Wireless Wireless-only Landline without Wireless Phoneless Landline with Wireless Wireless Only Landline without Wireless Phoneless Note: Adults are aged 18 and over, children are under age 18 Source: CDC/NCHS National Health Interview Survey Federal Communications Commission DA 15-1487 100 VIII. CONCLUSION 157. Promoting competition is a fundamental goal of the Commission’s policymaking. Competition has played and must continue to play an essential role in the mobile wireless industry – leading to lower prices and higher quality for American consumers, and producing innovation and investment in wireless networks, devices, and services. This Report analyzes competition in the mobile wireless industry pursuant to section 332(c)(1)(C) of the Communications Act and highlights several key trends. As with past reports, this Report examines various facets of the mobile wireless industry including market concentration, the conduct and rivalry of service providers, and competition in other segments of the mobile wireless ecosystem, including spectrum, backhaul, and handsets/devices, as well as consumer behavior. IX. PROCEDURAL MATTERS 158. This Eighteenth Report is issued pursuant to authority contained in Section 332(c)(1)(C) of the Communications Act of 1934, as amended, 47 U.S.C. § 332(c)(1)(C), and authority delegated to the Wireless Telecommunications Bureau under section 0.331 of the Commission’s rules, 47 C.F.R. § 0.331. 159. It is ORDERED that copies of this Eighteenth Report be sent to the appropriate committees and subcommittees of the United States House of Representatives and the United States Senate. FEDERAL COMMUNICATIONS COMMISSION Roger C. Sherman Chief, Wireless Telecommunications Bureau Federal Communications Commission DA 15-1487 101 APPENDIX I: MOBILE WIRELESS 3G OR BETTER COVERAGE BY NUMBER OF PROVIDERS470 Note: The percentages of population located in census blocks with a certain number of mobile service providers represent network coverage, which does not necessarily mean that they offered service to residents in the census block. In addition, we emphasize that a service provider reporting mobile wireless coverage in a particular census block may not provide coverage everywhere in the census block. For both these reasons, the number of service providers in a census block does not necessarily reflect the number of choices available to a particular individual or household, and does not purport to measure competition. In addition, calculations based on Mosaik data on coverage, while useful for measuring developments in mobile wireless coverage, have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. 470 We note that additional coverage maps showing mobile wireless digital coverage (CDMA, GSM/TDMA, and iDEN), nationwide mobile wireless coverage, and nationwide mobile LTE coverage are provided in (web) Appendix 1, available at https://www.fcc.gov/reports- research/reports?og_group_ref_target_id=1638&field_report_series_tid=1733&shs_term_node_tid_depth=All&=Apply. Federal Communications Commission DA 15-1487 102 APPENDIX II: COMPETITIVE DYNAMICS WITHIN THE INDUSTRY Connections, Net Adds, and Churn Table II.B.i Estimated Total Mobile Wireless Connections: 2002 – 2014 NRUF CTIA Year Connections (millions) Increase from previous year (millions) Connections Per 100 People Estimated Connections (millions) 2002 141.8 13.3 49 140.8 2003 160.6 18.8 54 158.7 2004 184.7 24.1 62 182.1 2005 213.0 28.3 71 207.9 2006 241.8 28.8 80 233.0 2007 263.0 21.2 86 255.4 2008 279.6 16.6 91 270.3 2009 290.7 11.1 94 285.6 2010 301.8 11.1 97 296.3 2011 317.3 15.5 101 316.0 2012 329.2 11.9 105 326.5 2013 339.2 10.0 108 335.7 2014 357.1 17.2 114 355.4 Source: Based on CTIA Wireless Industry Indices Year-End 2014, Table 6. NRUF 2002 – 2014. 2010 Census data (2010 nationwide population including the territories is 312,846,492). Table II.B.ii Quarterly Total Mobile Wireless Connections by Service Segment 2012 – 2nd Quarter 2015 Quarter Year Postpaid Prepaid Wholesale Connected Devices Total Connections 1Q12 215,466 69,133 13,955 24,502 323,056 2Q12 215,633 70,649 13,423 24,982 324,687 3Q12 216,129 71,112 13,567 25,836 326,644 4Q12 218,246 71,728 13,416 26,889 330,279 1Q13 217,887 73,007 16,847 28,233 335,974 2Q13 218,473 71,687 17,445 29,551 337,156 3Q13 221,142 71,906 17,881 30,932 341,862 4Q13 223,759 72,978 18,683 31,958 347,378 1Q14 225,580 74,827 17,738 33,661 351,807 2Q14 228,348 73,875 18,713 35,234 356,170 3Q14 231,572 73,774 20,210 38,462 364,017 4Q14 235,349 74,484 21,148 40,490 371,471 1Q15 237,409 74,606 22,236 41,961 376,213 2Q15 240,108 74,653 23,575 43,970 382,307 Source: UBS Investment Research. US Wireless 411, Version 51. US Wireless 411, Version 57. Federal Communications Commission DA 15-1487 103 Table II.B.iii Quarterly Net Adds in Mobile Wireless Connections by Service Segment (in thousands) 2010 – 2nd Quarter 2015 Quarter Year Postpaid Prepaid Wholesale Connected Devices Total Net Adds 1Q10 (6) 2,464 671 1,237 4,366 2Q10 813 749 483 1,421 3,466 3Q10 823 1,565 607 1,634 4,629 4Q10 895 2,633 39 1,831 5,398 1Q11 196 2,661 1,210 1,725 5,791 2Q11 787 1,093 1,017 1,452 4,349 3Q11 583 1,730 1,119 1,446 4,878 4Q11 1,304 1,998 1,506 76 4,884 1Q12 (147) 1,891 1,296 493 3,533 2Q12 784 414 568 480 2,246 3Q12 (405) 462 1,244 854 2,156 4Q12 2,177 603 (151) 1,053 3,682 1Q13 (3,872) 1,278 3,431 1,344 2,181 2Q13 263 (1,391) 598 1,318 788 3Q13 1,132 280 436 1,381 3,229 4Q13 2,492 1,069 802 1,026 5,389 1Q14 3,147 472 -945 1,703 4,378 2Q14 2,899 (1,029) 975 1,574 4,418 3Q14 3,064 686 1,497 1,967 7,214 4Q14 3,787 712 938 2,028 7,465 1Q15 2,003 187 1,088 1,471 4,748 2Q15 2,809 (21) 1,339 2,009 6,136 Source: 2010 data from the Sixteenth Competition Report, 28 FCC Rcd at 3836; UBS US Wireless 411 4Q11, at p. 10. 1Q12 - 2Q14 data from UBS US Wireless 411: Version 54; UBS US Wireless 411: Version 57. Federal Communications Commission DA 15-1487 104 Financial Indicators Table II.D.i Annualized Average Revenue Per Reported Subscriber Unit (ARPU): 1993 – 2014 Year Total Annual Service Revenue Percentage Change Average Reported Subscribers Average Monthly Revenue per Active Subscriber Unit 1993 $10,895,174,566 11,861,362 $76.55 1994 $14,229,921,264 30.6% 18,299,487 $64.80 1995 $19,081,239,000 34.1% 26,757,320 $59.43 1996 $23,634,971,000 23.9% 35,554,818 $55.40 1997 $27,485,632,936 16.3% 46,375,849 $49.39 1998 $33,133,174,978 20.6% 58,455,471 $47.23 1999 $40,018,489,104 20.8% 71,885,076 $46.39 2000 $52,466,019,720 31.1% 90,048,320 $48.55 2001 $65,316,235,000 24.5% 109,318,848 $49.79 2002 $76,508,187,000 17.1% 125,002,023 $51.00 2003 $87,624,093,000 14.5% 141,658,059 $51.55 2004 $102,121,210,043 16.5% 161,980,026 $52.54 2005 $113,538,220,438 11.2% 186,801,940 $50.65 2006 $125,456,824,884 10.5% 213,077,033 $49.07 2007 $138,869,303,958 10.7% 234,921,960 $49.26 2008 $148,084,169,893 6.6% 252,539,475 $48.87 2009 $152,551,853,953 3.0% 265,038,212 $47.97 2010 $159,929,646,977 4.9% 280,392,201 $47.53 2011 $169,767,314,353 6.2% 306,840,648 $46.11 2012 $185,013,934,995 9.0% 314,685,754 $48.99 2013 $189,192,811,836 2.3% 323,133,932 $48.79 2014 $187,848,477,106 (0.7%) 335,606,098 $46.64 Source: Based on CTIA Year-End 2014 Wireless Indices Report, Table 27. Federal Communications Commission DA 15-1487 105 Table II.D.ii Change in CPI, 1997 - 2014 Year CPI Wireless Telephone Services CPI Telephone Services CPI Land-line Telephone Services CPI Index Value Annual Change Index Value Annual Change Index Value Annual Change Index Value Annual Change 1997 100 100 100 - - 1998 101.6 95.1 100.7 - - 1999 103.8 2.2% 84.9 -10.7% 100.1 -0.6% - - 2000 107.3 3.4% 76.0 -10.5% 98.5 -1.6% - - 2001 110.3 2.8% 68.1 -10.4% 99.3 0.8% - - 2002 112.1 1.6% 67.4 -1.0% 99.7 0.4% - - 2003 114.6 2.3% 66.8 -0.9% 98.3 -1.4% - - 2004 117.7 2.7% 66.2 -0.9% 95.8 -2.5% - - 2005 121.7 3.4% 65.0 -1.8% 94.9 -0.9% - - 2006 125.6 3.2% 64.6 -0.6% 95.8 0.9% - - 2007 129.2 2.8% 64.4 -0.3% 98.2 2.6% - - 2008 134.1 3.8% 64.2 -0.2% 100.5 2.2% - - 2009 133.7 -0.4% 64.3 0.0% 102.4 1.9% - - 2010 135.9 1.6% 62.4 -2.9% 102.4 0.0% 101.6 - 2011 140.1 3.2% 60.1 -3.6% 101.2 -1.1% 103.3 1.7% 2012 143.0 2.1% 59.4 -1.2% 101.7 0.04% 106.1 2.7% 2013 145.2 1.5% 58.2 -1.6% 101.6 0.0% 109.3 3.0% 2014 147.4 1.6% 57.4 -2.1% 101.1 -0.4% 111.1 2.7% 1997 to 2014 35.6% -44.8% 0.8% 10.1% Source: Data from Bureau of Labor Statistics. All CPI figures were taken from BLS databases found on the BLS Internet site, available at http://www.bls.gov. Beginning in January 2010, the CPIs for local telephone service and long-distance telephone service were discontinued and replaced by a new CPI for land-line telephone services. Federal Communications Commission DA 15-1487 106 APPENDIX III: OVERALL WIRELESS INDUSTRY METRICS The tables and charts below are based on Commission estimates derived from census block analysis of Mosaik CoverageRight coverage maps, July 2015.471 We note that the percentage of the population located in census blocks with a certain number of service providers represents network coverage, which does not necessarily mean service is offered to residents in the census block. In addition, we emphasize that a service provider reporting coverage in a particular census block may not provide coverage everywhere in the census block. For both these reasons, the number of service providers in a census block does not necessarily reflect the number of choices available to a particular individual or household, and does not purport to measure competition. In addition, calculations based on Mosaik coverage data, while useful for measuring developments in mobile wireless coverage, have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Table III.A.i Estimated Mobile Wireless Coverage by Census Block Incl. Federal Land, July 2015 Number of Providers with Coverage in a Block Number of Blocks POPs Contained in Those Blocks % of Total U.S. POPs Square Miles Contained in Those Blocks % of Total U.S. Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles U.S. Total 11,155,486 312,471,327 100.0 3,802,067 100.0 6,821,187 100.0 1 or more 11,016,076 312,294,374 99.9 2,963,795 78.0 6,624,159 97.1 2 or more 10,746,142 311,349,672 99.6 2,611,973 68.7 6,253,679 91.7 3 or more 9,708,349 303,672,063 97.2 1,883,971 49.6 5,131,820 75.2 4 or more 8,150,122 286,467,076 91.7 1,161,936 30.6 3,773,219 55.3 5 or more 2,321,216 60,650,171 19.4 362,650 9.5 1,150,323 16.9 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 471 Population data are from the 2010 Census, and include the United States and Puerto Rico. Square miles also include the United States and Puerto Rico. Federal Communications Commission DA 15-1487 107 Table III.A.ii Estimated Mobile Wireless Coverage by Census Block Excl. Federal Land, July 2015 Number of Providers with Coverage in a Block Number of Blocks POPs Contained in Those Blocks % of Total U.S. POPs Square Miles Contained in Those Blocks % of Total U.S. Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles U.S. Total 10,449,282 307,208,959 100.0 2,664,706 100.0 5,893,270 100.0 1 or more 10,382,225 307,098,118 100.0 2,308,042 86.6 5,818,134 98.7 2 or more 10,201,965 306,332,361 99.7 2,139,846 80.3 5,608,895 95.2 3 or more 9,312,135 299,203,374 97.4 1,632,153 61.3 4,742,335 80.5 4 or more 7,890,214 282,721,581 92.0 1,058,820 39.7 3,579,215 60.7 5 or more 2,243,484 59,683,913 19.4 339,889 12.8 1,100,971 18.7 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. Chart III.A.i Estimated Mobile Wireless 3G or Better Coverage by Census Block, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. 5.0% 22.8% 41.3% 66.0% 76.2% 9.6% 45.7% 67.2% 89.9% 96.6% 12.6% 88.0% 95.5% 99.5% 99.9% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more 3G or Better Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Population % of U.S. Road Miles % of U.S. Square MilesSource: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 108 Table III.A.iii Estimated Mobile Wireless 3G or Better Coverage by Census Block Incl. Federal Land, July 2015 Number of Providers with Coverage in a Block Number of Blocks POPs Contained in Those Blocks % of Total U.S. POPs Square Miles Contained in Those Blocks % of Total U.S. Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles U.S. Total 11,155,486 312,471,327 100.0 3,802,067 100.0 6,821,187 100.0 1 or more 10,989,916 312,202,031 99.9 2,898,617 76.2 6,585,914 96.6 2 or more 10,662,014 311,014,320 99.5 2,510,098 66.0 6,133,402 89.9 3 or more 9,156,885 298,502,327 95.5 1,569,270 41.3 4,584,799 67.2 4 or more 7,269,188 275,130,421 88.0 865,111 22.8 3,120,038 45.7 5 or more 1,426,253 39,389,479 12.6 189,175 5.0 655,795 9.6 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Table III.A.iv Estimated Mobile Wireless 3G or Better Coverage by Census Block Excl. Federal Land, July 2015 Number of Providers with Coverage in a Block Number of Blocks POPs Contained in Those Blocks % of Total U.S. POPs Square Miles Contained in Those Blocks % of Total U.S. Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles U.S. Total 10,449,282 307,208,959 100.0 2,664,706 100.0 5,893,270 100.0 1 or more 10,366,375 307,042,064 99.9 2,280,745 85.6 5,799,540 98.4 2 or more 10,139,497 306,055,154 99.6 2,084,969 78.2 5,533,794 93.9 3 or more 8,824,450 294,291,060 95.8 1,401,983 52.6 4,298,999 72.9 4 or more 7,063,962 271,690,765 88.4 808,420 30.3 2,994,361 50.8 5 or more 1,380,956 38,743,888 12.6 179,259 6.7 630,889 10.7 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Federal Communications Commission DA 15-1487 109 Table III.A.v Estimated Mobile Wireless Network Coverage by Technology by Census Block, Mosaik, July 2015472 Technology POPs in Covered Blocks % of Total POPs Square Miles Contained in Those Blocks % of Total Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles CDMA 1xRTT 310,903,292 99.5 2,590,683 68.1 6,208,441 91.0 GPRS/EDGE 310,031,545 99.2 2,508,242 66.0 6,045,466 88.6 WCDMA/HSPA/HSPA+ 311,299,550 99.6 2,716,645 71.5 6,354,398 93.2 EV-DO/EV-DO Rev. A 310,594,906 99.4 2,487,400 65.4 6,093,222 89.3 Mobile WiMAX 105,546,452 33.8 43,978 1.2 419,921 6.2 LTE 311,115,606 99.6 2,573,689 67.7 6,185,589 90.7 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. Table III.A.vi Estimated Mobile Voice Coverage in Rural Areas by Census Block, July 2015 # of Providers with Coverage in a Block Number of Rural Census Blocks POPs Contained in Rural Census Blocks % of Total U.S. POPs Square Miles Contained in Those Blocks % of Total U.S. Square Miles Road Miles Contained in Those Blocks % of Total U.S. Road Miles Total for Rural U.S. 5,387,335 59,151,859 18.9 3,213,692 84.5 4,591,032 67.3 Total Rural U.S. POPs (%) Total Rural U.S. Square Miles (%) Total Rural U.S. Road Miles (%) 1 or more 5,174,914 58,775,012 99.4 2,288,106 71.2 4,277,960 93.2 2 or more 4,822,529 57,347,877 97.0 1,865,528 58.0 3,780,261 82.3 3 or more 3,798,592 49,682,553 84.0 1,212,368 37.7 2,704,800 58.9 4 or more 2,486,882 37,017,213 62.6 623,726 19.4 1,566,324 34.1 5 or more 910,191 14,345,261 24.3 190,415 5.9 517,884 11.3 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 472 The Commission has adopted flexible licensing policies and does not mandate any particular technology or network standard for mobile wireless licensees. Service providers choose their own network technologies and services and abide by certain technical parameters designed to avoid radiofrequency interference with adjacent licensees. Federal Communications Commission DA 15-1487 110 Table III.A.vii Estimated Mobile Voice Coverage in Non-Rural Areas by Census Block, July 2015 Number of Providers with Coverage in a Block Number of Non-Rural Census Blocks POPs Contained in Non-Rural Census Blocks % of Total U.S. POPs Square Miles in Those Blocks % of Total U.S. Square Miles Road Miles in Those Blocks % of Total U.S. Road Miles Total for Non- Rural U.S. 5,768,151 253,319,468 81.8 588,375 15.5 2,230,155 32.7 Total Non- Rural U.S. POPs (%) Total Non- Rural U.S. Square Miles (%) Total Non- Rural U.S. Road Miles (%) 1 or more 5,752,400 253,269,831 100.0 548,426 93.2 2,215,348 99.3 2 or more 5,721,720 253,040,079 99.9 525,354 89.3 2,184,864 98.0 3 or more 5,588,728 251,115,541 99.1 465,900 79.2 2,079,791 93.3 4 or more 5,176,461 242,007,399 95.5 356,433 60.6 1,813,178 81.3 5 or more 1,116,038 41,546,985 16.4 88,355 15.0 420,408 18.9 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. Chart III.A.ii Estimated Mobile Wireless Voice Coverage by Road Miles by Census Block in Rural vs. Non-Rural Areas, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. 11.3% 34.1% 58.9% 82.3% 93.2% 18.9% 81.3% 93.3% 98.0% 99.3% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile Wireless Voice CoverageN u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Non-Rural Road Miles % of U.S. Rural Road MilesSource: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 111 Table III.A.viii Estimated Mobile Wireless 3G or Better Coverage in Rural Areas by Census Block, July 2015 Total Number of Providers with Coverage in a Block Number of Rural Census Blocks POPs Contained in Rural Census Blocks % of Total Rural U.S. POPs Square Miles Contained in Those Blocks % of Total Rural U.S. Square Miles Road Miles Contained in Those Blocks % of Total Rural U.S. Road Miles 1 or more 5,231,510 58,903,268 99.6 2,345,458 73.0 4,366,137 95.1 2 or more 4,929,198 57,862,216 97.8 1,979,675 61.6 3,938,223 85.8 3 or more 3,588,316 47,515,484 80.3 1,113,264 34.6 2,525,449 55.0 4 or more 2,144,623 33,098,991 56.0 513,134 16.0 1,325,305 28.9 5 or more 654,006 10,392,502 17.6 127,143 4.0 361,759 7.9 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Table III.A.xix Estimated Mobile Wireless 3G or Better Coverage in Non-Rural Areas by Census Block, July 2015 Total Number of Providers with Coverage in a Block Number of Non-Rural Census Blocks POPs Contained in Non-Rural Census Blocks % of Total Non- Rural U.S. POPs Square Miles Contained in Those Blocks % of Total Non- Rural U.S. Square Miles Road Miles Contained in Those Blocks % of Total Non- Rural U.S. Road Miles 1 or more 5,758,406 253,298,763 100.0 553,159 94.0 2,219,778 99.5 2 or more 5,732,816 253,152,104 99.9 530,423 90.2 2,195,179 98.4 3 or more 5,568,569 250,986,843 99.1 456,006 77.5 2,059,350 92.3 4 or more 5,124,565 242,031,430 95.5 351,977 59.8 1,794,734 80.5 5 or more 772,247 28,996,977 11.4 62,032 10.5 294,037 13.2 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Federal Communications Commission DA 15-1487 112 Chart III.A.iii Estimated Mobile Wireless 3G or Better Coverage by Road Miles by Census Block, in Rural vs. Non-Rural Areas, July 2015 Source: Based on July 2015 Mosaik and 2010 Census data. It is important to note that the number of service providers in a census block represent network coverage only. Network coverage does not necessarily reflect the number of service providers from which any particular individual or household in a given area may choose. Coverage calculations based on Mosaik data, while useful for measuring developments in mobile coverage, have certain limitations that likely result in an overstatement of the extent of mobile coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. 7.9% 28.9% 55.0% 85.8% 95.1% 13.2% 80.5% 92.3% 98.4% 99.5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 or more 4 or more 3 or more 2 or more 1 or more Mobile Wireless 3G or Better Coverage (%) N u m b er o f S er v ic e P ro v id er s w it h C o v er a g e (J u ly 2 0 1 5 ) % of U.S. Non-Rural Road Miles % of U.S. Rural Road MilesSource: Mosaik Data, 2010 Census Federal Communications Commission DA 15-1487 113 APPENDIX IV: INPUT MARKET Table IV.B.i Year End Cell Site Counts by Service Provider, 2012 – 2014473 Source: Cell site counts for service providers are from UBS US Wireless 411: Version 57, Figure 63. The total industry-wide cell count is from CTIA Wireless Industry Indices Year- End 2014, at p. 103, Sept. 2015. 473 See CTIA Wireless Industry Indices Year-End 2014, at p. 101. Because multiple cell sites can be co-located in the same “tower” site, the reported cell sites should not be equated with “towers.” The reported cell sites include repeaters and other cell-extending devices (e.g., femtocells, or distributed antenna systems). See CTIA Wireless Industry Indices Year-End 2014, at pp. 101-2. Cell Sites 2012 2013 2014 1st Half 2015 Verizon Wireless 44,590 46,655 50,065 51,500 AT&T 56,900 61,800 71,768 67,383 Sprint 57,900 55,000 55,000 55,000 T-Mobile 51,104 63,879 61,079 58,651 NTELOS 1,429 1,444 1,453 1,443 US Cellular 8,028 6,975 6,220 6,223 Total by Top Seven Reported Service Providers 228,951 244,753 245,585 240,200 CTIA Reported Total Industry- wide Cell Sites 301,779 304,360 298,055 Federal Communications Commission DA 15-1487 114 APPENDIX V: PRICING The charts below show pricing changes for various plans across the four nationwide service providers, sourced from RBC Capital Markets, for the time period January 2014 through August 2015.474 474 See RBC Capital Markets, “Wireless Pricing Update, Wireless Telecommunications Service,” Jonathan Atkin and Brian Hyun (Aug. 26, 2015). $128 $122 $122 $122 $122 $102$102 $92 $70 $70 $70 $70 $97 $97 $97 $97 $97 $97 $143 $127 $112 $112 $107 $60 $70 $80 $90 $100 $110 $120 $130 $140 $150 Jan 14 Aug 14 Dec 14 Mar 15 May 15 Aug 15 Single Smartphone Plan (4 GB): Total Cost (Service Charge + EIP) AT&T Sprint T-Mobile Verizon Source: RBC Wireless Pricing Update, August 26, 2015 $185 $174 $174 $174 $174 $154 $174 $144 $144 $144 $144 $144 $154 $154 $154 $154 $154 $215 $184 $164 $120 $130 $140 $150 $160 $170 $180 $190 $200 $210 $220 Jan 14 Aug 14 Dec 14 Mar 15 May 15 Aug 15 Two Smartphones (min. 2 GB per line): Total Cost (Service Charge + EIP) AT&T Sprint T-Mobile Verizon Source: RBC Wireless Pricing Update, August 26, 2015 Federal Communications Commission DA 15-1487 115 $245 $226 $226 $226 $226 $226 6 $181 $171 $171 $181 $201 $201 $201 $201 $201 $201 $298 $251 $206 $196 $196 $150 $170 $190 $210 $230 $250 $270 $290 $310 Jan 14 Aug 14 Dec 14 Mar 15 May 15 Aug 15 Three Smartphones (min. 2 GB per line): Total Cost (Service Charge + EIP) AT&T Sprint T-Mobile Verizon Source: RBC Wireless Pricing Update, Aug. 26, 2015 $290 $268 $268 $268 $268 $268 $308 $208 $198 $198 $208 $248 $208 $208 $208 $228 $390 $248 $248 $248 $170 $200 $230 $260 $290 $320 $350 $380 $410 Jan 14 Aug 14 Dec 14 Mar 15 May 15 Aug 15 Four Smartphones (min. 2 GB per line): Total Cost (Service Charge + EIP) AT&T Sprint T-Mobile Verizon Source: RBC Wireless Pricing Update, Aug. 26, 2015 Federal Communications Commission DA 15-1487 116 APPENDIX VI: NON-PRICE RIVALRY Network Coverage The tables below are based on Commission estimates derived from census block analysis of Mosaik CoverageRight coverage maps, July 2015.475 We note that the percentage of the population located in census blocks with a certain number of service providers represents network coverage, which does not necessarily mean service is offered to residents in the census block. In addition, we emphasize that a service provider reporting coverage in a particular census block may not provide coverage everywhere in the census block. For both these reasons, the number of service providers in a census block does not necessarily reflect the number of choices available to a particular individual or household, and does not purport to measure competition. In addition, calculations based on Mosaik coverage data, while useful for measuring developments in mobile wireless coverage, have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Table VI.B.i Estimated Mobile Wireless Coverage in the U.S. by Service Provider, July 2015 Service Provider Number of Blocks POPS Contained in Those Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 11,155,486 312,471,327 100.0 100.0 100.0 AT&T 10,646,897 310,278,040 99.3 68.4 90.5 Verizon Wireless 10,300,641 304,170,059 97.3 63.7 87.0 Sprint 7,526,872 280,152,856 89.7 24.0 47.3 T-Mobile 8,183,337 286,265,928 91.6 34.0 58.0 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Table VI.B.ii Estimated Mobile Wireless 3G or Better Coverage in the U.S. by Service Provider, July 2015 Service Provider Number of Blocks POPS Contained in Those Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 11,155,486 312,471,327 100.0 100.0 100.0 AT&T 10,616,327 310,136,594 99.3 67.5 90.0 Verizon Wireless 10,287,446 304,107,377 97.3 63.1 86.7 Sprint 7,526,872 280,152,856 89.7 24.0 47.3 T-Mobile 7,077,781 272,483,385 87.2 23.8 45.7 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. 475 Population data are from the 2010 Census, and include the United States and Puerto Rico. Square miles also include the United States and Puerto Rico. Federal Communications Commission DA 15-1487 117 Table VI.B.iii Estimated LTE Broadband Coverage in the U.S. by Service Provider, July 2015 Service Provider Number of Blocks POPS Contained in Those Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 11,155,486 312,471,327 100.0 100.0 100.0 AT&T 9,515,982 302,070,044 96.7 46.4 72.2 Verizon Wireless 10,159,559 303,212,253 97.0 60.9 84.6 Sprint 6,637,924 265,642,199 85.0 17.7 38.6 T-Mobile 6,672,986 266,234,619 85.2 20.1 41.0 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Table VI.B.iv Estimated Mobile Wireless Voice Coverage in Rural Areas by Service Provider, July 2015 Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,387,335 59,151,859 18.9 84.5 67.3 Provider Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total Rural U.S. POPs % of Total Rural U.S. Square Miles % of Total Rural U.S. Road Miles AT&T 4,398,674 54,048,800 91.4 51.2 73.3 Verizon Wireless 4,501,096 53,715,705 90.8 55.7 78.3 Sprint 2,096,363 33,581,916 56.8 14.6 26.9 T-Mobile 2,651,166 37,784,502 63.9 24.6 40.1 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Federal Communications Commission DA 15-1487 118 Table VI.B.v Estimated Mobile Wireless Voice Coverage in Non-Rural Areas by Service Provider, July 2015 Number of Non-Rural Census Blocks POPS Contained in Non-Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,768,151 253,319,468 81.8 15.5 32.7 Provider Number of Non-Rural Census Blocks POPS Contained in Non-Rural Census Blocks % of Total Non-Rural U.S. POPs % of Total Non-Rural U.S. Square Miles % of Total Non-Rural U.S. Road Miles AT&T 5,719,061 252,918,393 99.8 89.8 97.9 Verizon Wireless 5,603,401 248,298,419 98.0 86.6 95.8 Sprint 5,112,435 239,696,324 94.6 59.2 78.9 T-Mobile 5,374,070 246,561,147 97.3 73.7 88.8 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Table VI.B.vi Estimated Mobile 3G or Better Coverage in Rural Areas by Service Provider, July 2015 Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,387,335 59,151,859 18.9 84.5 67.3 Provider Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total Rural U.S. POPs % of Total Rural U.S. Square Miles % of Total Rural U.S. Road Miles AT&T 4,869,669 56,940,755 96.3 63.1 85.7 Verizon Wireless 4,651,740 55,050,346 93.1 58.6 81.7 Sprint 2,246,358 35,567,035 60.1 16.2 29.5 T-Mobile 1,948,906 30,772,737 52.0 16.3 27.9 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Federal Communications Commission DA 15-1487 119 Table VI.B.vii Estimated Mobile 3G or Better Coverage in Non-Rural Areas by Service Provider, July 2015 Number of Non-Rural Census Blocks POPS Contained in Non-Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,768,151 253,319,468 81.8 15.5 32.7 Provider Number of Non-Rural Census Blocks POPS Contained in Non-Rural Census Blocks % of Total Non-Rural U.S. POPs % of Total Non-Rural U.S. Square Miles % of Total Non-Rural U.S. Road Miles AT&T 5,746,658 253,195,839 100.0 91.4 98.9 Verizon Wireless 5,635,706 249,057,031 98.3 88.2 96.8 Sprint 5,280,514 244,585,821 96.6 66.4 83.9 T-Mobile 5,128,875 241,710,648 95.4 64.7 82.3 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. The 3G and 4G technologies included are the following: EVDO, EVDO Rev A, WCDMA/HSPA, HSPA+, LTE, and mobile WiMAX. Table VI.B.viii Estimated LTE Broadband Coverage in Rural Areas by Service Provider, July 2015 Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,387,335 59,151,859 18.9 84.5 67.3 Provider Number of Rural Census Blocks POPS Contained in Rural Census Blocks % of Total Rural U.S. POPs % of Total Rural U.S. Square Miles % of Total Rural U.S. Road Miles AT&T 3,887,034 50,348,879 85.1 39.8 61.4 Verizon Wireless 4,546,276 54,451,645 92.1 56.2 79.0 Sprint 1,660,665 28,469,487 48.1 10.4 20.2 T-Mobile 1,677,041 28,004,770 47.3 12.7 22.6 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Federal Communications Commission DA 15-1487 120 Table VI.B.ix Estimated LTE Broadband Coverage in Non-Rural Areas by Service Provider, July 2015 Number of Non-Rural Census Blocks POPS Contained in Non- Rural Census Blocks % of Total U.S. POPs % of Total U.S. Square Miles % of Total U.S. Road Miles Total for U.S. 5,768,151 253,319,468 81.8 15.5 32.7 Service Provider Number of Non-Rural Census Blocks with LTE Coverage POPS Contained in Non- Rural Census Blocks with LTE Coverage % of Total Non- Rural U.S. POPs with LTE Coverage % of Total Non- Rural U.S. Square Miles with LTE Coverage % of Total Non- Rural U.S. Road Miles with LTE Coverage AT&T 5,628,948 251,721,165 99.4 82.1 94.5 Verizon Wireless 5,613,283 248,760,608 98.2 86.6 96.0 Sprint 4,977,259 237,172,712 93.6 57.4 76.6 T-Mobile 4,995,945 238,229,849 94.0 60.4 78.9 Source: July 2015 Mosaik and 2010 Census. It is important to note that the data underlying these estimates measure mobile network coverage only. Network coverage does not necessarily reflect whether a service provider is indeed affirmatively offering service to residents in all these geographical areas. In addition, the Mosaik data as well as the coverage calculation methodology have certain limitations that likely result in an overstatement of the extent of mobile wireless coverage. Federal Communications Commission DA 15-1487 121 Quality of Service Ookla: An in-depth discussion of the Ookla speed test is available in the Seventeenth Report.476 In this Report, we present mobile wireless upload and download speeds within the United States for the second half of 2014 through the first half of 2015.477 Commission staff estimated the total value of each parameter for all service providers using the specific technology combination in the applicable geographic region. For each geographic and technological subgroup, the total mean value is calculated to be the mean of all the state means, and the total median is the median of all the state medians. Table VI.C.i Ookla Speed Test - Estimated Download Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests (’000s) Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests (’000s) Mean Down load Speed (Mbps) Median Download Speed (Mbps) Number of tests (’000s) Verizon Wireless 15.26 11.86 4,115 15.62 12.18 7,082 16.11 12.65 2,968 AT&T 10.66 7.29 5,465 10.97 7.52 9,191 11.44 7.86 3,726 Sprint 6.83 4.68 3,627 7.24 4.93 6,044 7.89 5.35 2,417 T-Mobile 12.90 10.23 6,207 13.50 10.66 11,302 14.19 11.04 5,093 Total 6.29 5.00 19,926 6.65 5.43 34,487 7.38 5.94 14,558 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Table VI.C.ii Ookla Speed Test - Estimated LTE Download Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests (’000s) Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests (’000s) Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests (’000s) Verizon Wireless 16.18 12.77 3,880 16.39 12.95 6,747 16.67 13.23 2,867 AT&T 13.33 10.38 4,122 13.29 10.18 7,223 13.27 9.93 3,101 Sprint 9.26 7.47 2,893 9.55 7.46 5,024 10.00 7.73 2,130 T-Mobile 18.45 16.79 4,368 17.54 15.23 8,518 17.31 14.39 4,150 Total 11.63 10.20 15,563 11.48 10.01 28,060 11.20 9.51 12,495 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. 476 See Seventeenth Report, 29 FCC Rcd at 15465-66, Appendix VI. ¶¶ 1-6. 477 The provider specific mean, median, and count values shown below were provided to the Commission by Ookla, who estimated these parameters by service provider, at the state and nationwide level, for all technologies and for LTE only. Federal Communications Commission DA 15-1487 122 Table VI.C.iii Ookla Speed Test - Estimated Upload Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Verizon Wireless 6.00 4.57 4,115 6.10 4.56 7,082 6.25 4.58 2,968 AT&T 4.18 2.75 5,465 4.31 2.96 9,191 4.51 3.27 3,726 Sprint 2.93 2.11 3,627 3.08 2.23 6,044 3.31 2.45 2,417 T-Mobile 5.45 4.00 6,207 5.88 4.58 11,302 6.38 5.27 5,093 Total 2.68 2.07 19,926 2.88 2.32 34,487 3.25 2.63 14,558 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Table VI.C.iv Ookla Speed Test - Estimated LTE Upload Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Verizon Wireless 6.34 5.00 3,880 6.39 4.91 6,747 6.46 4.83 2,867 AT&T 5.38 4.50 4,122 5.37 4.48 7,223 5.36 4.45 3,101 Sprint 3.97 3.60 2,893 4.02 3.55 5,024 4.21 3.64 2,130 T-Mobile 8.13 7.68 4,368 8.02 7.45 8,518 8.04 7.34 4,150 Total 5.15 4.60 15,563 5.27 4.72 28,060 5.18 4.56 12,495 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Federal Communications Commission DA 15-1487 123 Table VI.C.v Ookla Speed Test - Estimated Download Speeds by Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests (’000s) Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests (’000s) Mean Down load speed (Mbps) Median Download speed (Mbps) Number of tests (’000s) Verizon Wireless 16.85 11.45 696 18.34 12.64 1,151 20.51 14.65 456 AT&T 11.89 7.56 966 12.24 7.78 1,650 12.70 8.07 684 Sprint 6.65 4.12 578 7.34 4.53 923 8.50 5.26 344 T-Mobile 14.77 11.27 1,596 14.77 10.70 2,930 14.77 9.98 1,334 Total 9.11 8.01 3,912 8.50 7.53 6,774 7.42 6.15 2,861 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Table VI.C.vi Ookla Speed Test - Estimated LTE Download Speeds by Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests (’000s) Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests (’000s) Mean Down load speed (Mbps) Median Download speed (Mbps) Number of tests (’000s) Verizon Wireless 17.48 12.12 664 18.94 13.31 1,107 21.03 15.27 442 AT&T 13.95 9.87 754 14.05 9.74 1,332 14.17 9.57 579 Sprint 8.69 6.32 440 9.19 6.49 736 9.93 6.78 295 T-Mobile 18.50 15.96 1,146 17.81 14.60 2,247 17.06 12.86 1,101 Total 13.86 12.60 3,052 14.48 13.18 5,500 11.09 9.36 2,448 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Federal Communications Commission DA 15-1487 124 Table VI.C.vii Ookla Speed Test - Estimated Upload Speeds by Top 4 Nationwide Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Verizon Wireless 7.12 4.70 696 7.62 4.88 1,151 8.36 5.17 456 AT&T 5.00 2.81 966 5.12 3.07 1,650 5.28 3.40 684 Sprint 2.58 1.14 578 2.81 1.32 923 3.20 1.60 344 T-Mobile 7.85 5.69 1,596 8.16 6.21 2,930 8.52 6.80 1,334 Total 4.38 3.79 3,912 3.90 3.35 6,774 3.22 2.57 2,861 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Table VI.C.viii Ookla Speed Test - Estimated LTE Upload Speeds by Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Mean Upload speed (Mbps) Median Upload speed (Mbps) Number of tests (’000s) Verizon Wireless 7.38 4.99 664 7.87 5.15 1,107 8.57 5.42 442 AT&T 6.06 4.44 754 6.06 4.47 1,332 6.05 4.51 579 Sprint 3.28 2.32 440 3.44 2.32 736 3.68 2.32 295 T-Mobile 10.35 9.11 1,146 10.28 9.01 2,247 10.20 8.91 1,101 Total 6.95 6.44 3,052 6.94 6.37 5,500 5.11 4.43 2,448 Source: Ookla Internet performance data, © 2015 Ookla, LLC. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers, as well as any other wireless service providers that appear in the Ookla dataset. Federal Communications Commission DA 15-1487 125 FCC: An in-depth discussion of the FCC Speed test is available in the Seventeenth Report.478 In this Report, we present mobile wireless upload and download speeds within the United States for the second half of 2014 through the first half of 2015. These values are estimated using all technologies, and also for LTE only.479 We note that throughput is originally measured in bytes/sec, and the conversion to Mbps is [1 Mbps=8*(10^-6) bytes/sec]. Roaming observations are included in the mean and median speed estimates, in order to mimic actual customer experience. Total mean and median download speeds are calculated using data for the four nationwide providers, as well as any other wireless service providers that appear in the FCC dataset. The data collected by the Measuring Broadband America Program's FCC Speed Test App for Android and iOS Speed data were used in this Report. These data included the test results for download, upload, and latency as well as information about the handset and cellular environment such as the location of the test, operating system of the handset, the WiFi or 3G/4G cellular technology used for the test, the bearer channels, and certain other parameters helpful for the characterization of mobile performance. Data extracts for the appropriate time periods reviewed in this Report were culled from the main database collection selecting only observations that included valid GPS location data in the United States and were executed on the smartphone’s cellular connection. So, for example, if a test was executed on a WiFi connection in any portion of a download test, that observation was not included. In addition, only observations with valid and consistent network operator names were included in the data. Duplicate measurements were dropped. 478 See Seventeenth Report, 29 FCC Rcd at 15467, Appendix VI. ¶¶ 7-9. 479 Throughput speeds were converted from bytes/sec to Mbps using a conversion of [1 Mbps=8*(10^-6) bytes/sec]. The throughput speed for failed tests was replaced with a value of zero, and the top 1% of speed observations were trimmed from the dataset, by service provider and separately for each time period. Roaming observations were included in the calculations in order to replicate customer experience; however, these observations did not make up a significant portion of the data. Only observations within the United States were included in the analysis. All tests performed over Wi-Fi were dropped, and only mobile observations were included in the analysis. Federal Communications Commission DA 15-1487 126 Table VI.C.ix FCC Speed Test - Estimated Download Speeds for All Technologies By Service Provider Including Discount Brands, Nationwide480 Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load speed (Mbps) Median Down load speed (Mbps) Number of tests Verizon 14.08 8.69 76,377 14.52 9.23 148,551 14.97 9.83 72,174 AT&T (includes Cricket) 8.76 5.14 46,807 8.62 5.05 100,313 8.51 4.96 53,507 Sprint (includes Virgin, Boost) 3.88 1.32 49,221 4.45 1.59 89,853 5.16 1.93 40,632 T-Mobile (includes Metro PCS) 12.21 8.24 58,911 12.91 8.26 121,225 13.58 8.27 62,314 Total 9.88 5.06 251,943 10.35 5.37 501,105 10.83 5.69 249,163 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands as did the Seventeenth Report. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – June 2015. 480 Consistent with the methodology used in previous reports, estimated download speed data includes any discount brands offered by a service provider. We recognize that some discount brands may have capped LTE download speeds, or employ different network management practices. See, e.g., CricketWireless website, “Mobile Broadband Information,” available at https://www.cricketwireless.com/legal-info/mobile-broadband-information.html; Virgin Mobile website, available at http://coverage.sprint.com/coverageDescVMU.html. Federal Communications Commission DA 15-1487 127 Table VI.C.x FCC Speed Test - Estimated LTE Download Speeds by Service Provider’s Flagship Brand, Nationwide481 Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Verizon 16.13 10.97 63,436 16.34 11.24 110,549 16.63 11.66 47,114 AT&T 12.13 8.52 23,913 11.78 8.06 44,818 11.38 7.53 20,905 Sprint 7.13 5.00 23,701 7.47 5.02 41,430 7.94 5.05 17,729 T-Mobile 17.97 14.91 31,586 18.02 14.33 59,594 18.07 13.58 28,008 Total 13.98 9.44 154,279 14.05 9.29 280,373 14.13 9.07 126,096 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data excludes discount brands. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – May 2015. Table VI.C.xi FCC Speed Test - Estimated LTE Download Speeds by Service Provider Including Discount Brands, Nationwide482 Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Verizon 16.13 10.97 63,436 16.34 11.24 110,549 16.63 11.66 47,114 AT&T (includes Cricket) 11.26 7.69 26,596 10.69 7.31 51,604 10.07 6.84 25,009 Sprint (includes Virgin, Boost) 7.11 4.98 24,156 7.43 4.98 42,435 7.86 4.99 18,280 T-Mobile (includes MetroPCS) 17.65 14.57 33,195 17.73 13.99 63,309 17.82 13.27 30,114 Total 13.98 9.44 154,280 14.05 9.28 280,372 14.14 9.07 126,096 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – May 2015. 481 Estimated download speed data excludes any discount brands offered by a service provider. 482 See n.480 supra. Federal Communications Commission DA 15-1487 128 Table VI.C.xii FCC Speed Test - Estimated Upload Speeds for All Technologies by Service Provider Including Discount Brands, Nationwide483 Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.05 2.68 72,013 4.97 2.96 140,996 5.94 3.28 68,984 AT&T (includes Cricket) 3.01 1.66 41,792 3.59 1.87 89,677 4.11 2.09 47,885 Sprint (includes Virgin, Boost) 1.58 0.73 45,926 1.88 0.80 83,495 2.27 0.88 37,570 T-Mobile (includes MetroPCS) 4.25 2.33 53,925 5.64 2.91 108,278 7.03 4.02 54,353 Total 3.22 1.53 232,962 4.05 1.79 460,855 4.91 2.16 227,895 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – June 2015. Table VI.C.xiii FCC Speed Test - Estimated LTE Upload Speeds by Service Provider’s Flagship Brand, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.63 3.54 60,286 5.59 3.82 105,888 6.86 4.31 45,603 AT&T 4.10 3.47 22,150 4.73 3.67 41,579 5.45 4.01 19,429 Sprint 2.75 2.22 21,725 3.08 2.36 38,129 3.55 2.58 16,405 T-Mobile 6.86 5.63 28,789 8.47 6.50 53,963 10.31 8.64 25,175 Total 4.67 3.68 143,477 5.66 4.05 261,224 6.87 4.71 117,750 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data excludes discount brands. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – May 2015. 483 See n.480 supra. Federal Communications Commission DA 15-1487 129 Table VI.C.xiv FCC Speed Test - Estimated LTE Upload Speeds by Service Provider Including Discount Brands, Nationwide484 Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.63 3.54 60,286 5.59 3.82 105,888 6.86 4.31 45,603 AT&T (includes Cricket) 4.07 3.36 24,453 4.84 3.73 47,613 5.67 4.28 23,161 Sprint (includes Virgin, Boost) 2.75 2.22 22,129 3.07 2.33 39,051 3.51 2.52 16,923 T-Mobile (includes MetroPCS) 6.80 5.58 30,179 8.44 6.46 56,993 10.29 8.63 26,815 Total 4.67 3.68 143,478 5.66 4.05 261,225 6.87 4.71 117,751 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – May 2015. 484 See id. Federal Communications Commission DA 15-1487 130 Table VI.C.xv FCC Speed Test - Estimated Download Speeds for All Technologies by Service Provider Including Discount Brands, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Verizon 13.51 7.31 10,116 14.66 8.12 20,293 15.80 9.08 10,177 AT&T (includes Cricket) 9.07 5.39 5,864 9.12 5.20 12,842 9.17 5.05 6,978 Sprint (includes Virgin, Boost) 3.05 1.00 6,312 3.61 1.16 11,068 4.40 1.47 4,756 T-Mobile (includes MetroPCS) 12.43 8.59 10,396 12.80 8.06 23,013 13.11 7.49 12,617 Total 9.97 4.96 34,477 10.81 5.29 70,339 11.62 5.65 35,862 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – June 2015. Table VI.C.xvi FCC Speed Test - Estimated LTE Download Speeds by Service Provider’s Flagship Brand, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Verizon 14.72 9.02 8,466 15.68 9.61 15,206 16.90 10.52 6,740 AT&T 11.63 7.59 2,963 11.23 6.98 5,856 10.84 6.51 2,894 Sprint 6.59 4.69 2,522 7.07 4.61 4,335 7.84 4.53 1,813 T-Mobile 17.80 14.90 5,511 17.60 13.92 11,084 17.41 12.92 5,573 Total 13.69 8.94 20,730 14.07 8.81 39,193 14.52 8.64 18,465 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data excludes discount brands. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – May 2015. Federal Communications Commission DA 15-1487 131 Table VI.C.xvii FCC Speed Test - Estimated LTE Download Speeds by Service Provider Including Discount Brands, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of tests Verizon 14.72 9.02 8,466 15.68 9.61 15,206 16.90 10.52 6,740 AT&T (includes Cricket) 11.42 7.37 3,051 10.70 6.64 6,288 10.04 6.09 3,238 Sprint (includes Virgin, Boost) 6.58 4.68 2,566 7.03 4.59 4,418 7.78 4.46 1,853 T-Mobile (includes MetroPCS) 16.80 13.67 6,175 16.62 12.84 12,498 16.45 11.80 6,324 Total 13.69 8.94 20,730 14.07 8.81 39,193 14.52 8.64 18,466 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all LTE service providers, not only the four nationwide service providers. July 2014 – June 2015. Table VI.C.xviii FCC Speed Test - Estimated Upload Speeds for All Technologies by Service Provider Including Discount Brands, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.03 2.48 9,348 5.25 2.67 18,968 6.44 2.91 9,620 AT&T (includes Cricket) 3.26 1.39 5,165 3.72 1.66 11,244 4.12 1.91 6,079 Sprint (includes Virgin, Boost) 1.29 0.65 5,858 1.44 0.67 9,999 1.67 0.72 4,142 T-Mobile (includes MetroPCS) 5.06 2.61 9,354 6.62 3.83 19,463 8.06 5.33 10,109 Total 3.55 1.40 31,339 4.62 1.75 62,487 5.71 2.26 31,149 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – May 2015. Federal Communications Commission DA 15-1487 132 Table VI.C.xix FCC Speed Test - Estimated Upload LTE Speeds by Service Provider’s Flagship Brand, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.47 3.09 7,959 5.64 3.21 14,428 7.09 3.40 6,469 AT&T 4.47 3.43 2,771 4.98 3.55 5,484 5.52 3.67 2,714 Sprint 2.52 1.74 2,254 2.62 1.55 3,736 2.81 1.31 1,483 T-Mobile 8.05 6.52 4,918 9.88 7.45 9,760 11.75 10.13 4,843 Total 5.17 3.89 18,985 6.42 4.24 35,485 7.87 4.82 16,503 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data excludes discount brands. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – May 2015. Table VI.C.xx FCC Speed Test - Estimated Upload LTE Speeds by Service Provider Including Discount Brands, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of tests Verizon 4.47 3.09 7,959 5.64 3.21 14,428 7.09 3.40 6,469 AT&T (includes Cricket) 4.46 3.42 2,841 5.00 3.54 5,788 5.52 3.66 2,948 Sprint (includes Virgin, Boost) 2.52 1.74 2,290 2.61 1.55 3,805 2.79 1.30 1,515 T-Mobile (includes MetroPCS) 7.77 6.34 5,506 9.66 7.24 10,823 11.62 9.90 5,318 Total 5.17 3.89 18,984 6.42 4.24 35,484 7.87 4.82 16,503 Source: Data from FCC Measuring Mobile Broadband America data. Observations include failed tests. Provider speed data includes discount brands, as did the Seventeenth Report. Total speeds are evaluated using data for all service providers, not only the four nationwide service providers. July 2014 – May 2015. Federal Communications Commission DA 15-1487 133 Rootmetrics. An in-depth discussion of the Root Metrics dataset is available in the Seventeenth Report.485 In this Report, we present mobile wireless upload and download speeds and indices within the United States for the second half of 2014 through the first half of 2015.486 Table VI.C.xxi Root Metrics National Speed Index Data, 2nd Half 2014, 1st Half 2015 2nd Half 2014 1st Half 2015 Service Provider Speed Index Data Performance Text Performance Speed Index Data Performance Text Performance Verizon Wireless 89.0 94.5 92.4 91.9 96.6 95.2 AT&T 85.5 91.5 93.2 88.5 94.1 95.5 Sprint 71.0 81.4 92.7 75.8 85.0 95.0 T-Mobile 79.1 81.9 89.7 85.1 87.0 90.5 Source: RootMetrics RootScore Report Data, 2nd half 2014, 1st half 2015. Table VI.C.xxii RootMetrics California Speed Index Data, 2nd Half 2014, 1st Half 2015 2nd Half 2014 1st Half 2015 Service Provider Speed Index Data Performance Text Performance Speed Index Data Performance Text Performance Verizon Wireless 90.9 95.2 91.8 93.0 97.2 95.4 AT&T 89.1 93.5 92.2 90.9 95.2 96.6 Sprint 67.4 77.9 91.6 77.0 86.4 95.2 T-Mobile 86.5 88.5 90.6 89.6 93.4 94.9 Source: RootMetrics RootScore Report Data, 2nd half 2014, 1st half 2015. 485 See Seventeenth Report, 29 FCC Rcd at 15467, Appendix VI. ¶¶ 10-11. 486 The speed estimates are based on median download and upload speeds by state and by provider, which RootMetrics provided to the Commission. To estimate “mean” nationwide speeds, the mean of the median values across all states was calculated, by provider and overall. Nationwide median speeds represent the median speed across all states, by service provider and overall. This was done separately for each time period. Federal Communications Commission DA 15-1487 134 Table VI.C.xxiii Root Metrics Speed Test - Estimated LTE Download Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load speed (Mbps) Number of tests Verizon Wireless 14.94 14.92 205,508 15.41 15.30 471,655 15.87 15.68 266,147 AT&T 9.98 9.74 206,082 9.88 9.83 472,202 9.77 9.93 266,120 Sprint 3.80 3.94 205,637 4.30 4.50 471,605 4.80 5.06 265,968 T-Mobile 9.29 9.37 205,397 9.47 9.43 471,186 9.65 9.49 265,789 Total 9.50 9.49 822,624 9.76 9.77 1,886,648 10.02 10.04 1,064,024 Source: RootMetrics Data, 2015. © Rootmetrics. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers. Table VI.C.xxiv Root Metrics Speed Test - Estimated LTE Download Speeds by Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Down load Speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Down load speed (Mbps) Number of tests Mean Down load Speed (Mbps) Median Download speed (Mbps) Number of tests Verizon Wireless 16.91 16.91 19,197 19.17 19.17 56,528 21.44 21.44 37,331 AT&T 12.10 12.10 19,205 12.33 12.33 56,493 12.56 12.56 37,288 Sprint 3.05 3.05 19,180 4.80 4.80 56,527 6.56 6.56 37,347 T-Mobile 12.35 12.35 19,194 12.75 12.75 56,404 13.15 13.15 37,210 Total 11.10 11.10 76,776 12.27 12.27 225,952 13.43 13.43 149,176 Source: RootMetrics Data, 2015. © Rootmetrics. All rights reserved. Total mean and median download speeds are calculated using data for the four nationwide service providers. Federal Communications Commission DA 15-1487 135 Table VI.C.xxv Root Metrics Speed Test - Estimated LTE Upload Speeds by Service Provider, Nationwide Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Verizon Wireless 8.16 7.91 205,153 8.56 8.09 471,740 8.97 8.27 266,587 AT&T 5.61 4.89 206,468 5.54 4.82 473,051 5.46 4.74 266,583 Sprint 2.13 2.32 206,023 2.35 2.43 472,437 2.58 2.54 266,414 T-Mobile 6.21 6.55 205,734 6.71 6.72 471,955 7.20 6.90 266,221 Total 5.52 5.42 823,378 5.79 5.51 1,889,183 6.05 5.61 1,065,805 Source: RootMetrics Data, 2015. © Rootmetrics. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers. Table VI.C.xxvi Root Metrics Speed Test - Estimated LTE Upload Speeds by Service Provider, California Only Service Provider 2H2014 2H2014 - 1H2015 1H2015 Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Mean Upload Speed (Mbps) Median Upload speed (Mbps) Number of tests Verizon Wireless 10.05 10.05 19,209 12.68 12.68 56,579 15.31 15.31 37,370 AT&T 7.80 7.80 19,220 7.84 7.84 56,554 7.88 7.88 37,334 Sprint 1.34 1.34 19,194 2.35 2.35 56,592 3.37 3.37 37,398 T-Mobile 10.65 10.65 19,206 12.95 12.95 56,450 15.25 15.25 37,244 Total 7.46 7.46 76,829 8.96 8.96 226,175 10.45 10.45 149,346 Source: RootMetrics Data, 2015. © Rootmetrics. All rights reserved. Total mean and median upload speeds are calculated using data for the four nationwide service providers. Federal Communications Commission DA 15-1487 136 CalSPEED. An in-depth discussion of the CalSPEED dataset is available in the Seventeenth Report.487 In this Report, we present mobile wireless upload and download speeds within California for the fall of 2014 and the spring of 2015.488 We also present charts summarizing mobile wireless upload and download speeds and latency within the United States for the spring of 2012 through the spring of 2015.489 Table VI.C.xxvii CalSPEED - Estimated Download Speeds by Service Provider, California Only Service Provider Fall 2014 Fall 2014 and Spring 2015 Spring 2015 Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of Tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of Tests Mean Down load Speed (Mbps) Median Down load Speed (Mbps) Number of Tests Verizon Wireless 12.51 10.71 1,865 11.75 10.71 3,515 10.99 10.71 1,651 AT&T 9.72 7.11 1,647 8.91 6.69 3,358 8.19 6.30 1,711 Sprint 3.17 0.96 1,234 3.57 1.21 2,586 3.95 1.48 1,353 T-Mobile 8.43 6.22 1,344 8.50 7.03 2,354 8.67 7.89 1,010 Total 8.96 5.88 6,090 8.51 6.01 11,813 8.08 6.16 5,725 Source: The estimated speeds are based on the CalSPEED data. The top 1% of speed values were dropped, by provider and time period. Fall 2014 tests were taken between the dates of 10/1/2014 and 11/21/2014. Spring 2015 tests were taken between the dates of 5/1/2015 and 6/15/2015. Total mean and median download speeds are calculated using data for the four nationwide service providers. 487 See Seventeenth Report, 29 FCC Rcd at 15469-70, Appendix VI. ¶¶ 12-16. 488 These values are estimated using all technologies, and also for LTE only. The throughput speed was replaced with a value of zero for certain test errors, which correspond to the method used by CPUC. Tests were not included if they were quit by the user, if the test was outside of the service area, or if the testing device was not a smartphone. Finally, results from each site and for each provider were averaged across all east coast and west coast servers, and the top 1% of resulting speed observations were trimmed from the dataset, by provider and separately for each time period. This is a surveyed test and not crowdsourced, and therefore some of the cleaning criteria may be different from the other speed tests. 489 We note that while CPUC performed tests using both commercially available, Android-based smartphones as well as tablets, the charts in this Appendix reflect measurements taken using phones only. Federal Communications Commission DA 15-1487 137 Table VI.C.xxviii CalSPEED - Estimated Upload Speeds by Service Provider, California Only Service Provider Fall 2014 Fall 2014 and Spring 2015 Spring 2015 Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of Tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of Tests Mean Upload Speed (Mbps) Median Upload Speed (Mbps) Number of Tests Verizon Wireless 6.38 6.33 1,865 6.53 6.51 3,515 6.70 6.69 1,651 AT&T 5.63 5.17 1,647 5.50 5.02 3,358 5.38 4.76 1,711 Sprint 2.12 0.87 1,234 2.29 0.95 2,586 2.46 1.06 1,353 T-Mobile 5.10 4.42 1,344 5.52 5.45 2,354 6.07 6.97 1,010 Total 5.03 4.02 6,090 5.11 4.13 11,813 5.19 4.23 5,725 Source: The estimated speeds are based on the CalSPEED data. The top 1% of speed values were dropped, by provider and time period. Fall 2014 tests were taken between the dates of 10/1/2014 and 11/21/2014. Spring 2015 tests were taken between the dates of 5/1/2015 and 6/15/2015. Total mean and median upload speeds are calculated using data for the four nationwide service providers. Table VI.C.xxix CalSPEED - Estimated LTE Download Speeds by Service Provider, California Only Service Provider Fall 2014 Fall 2014 and Spring 2015 Spring 2015 Mean LTE Down load Speed (Mbps) Median LTE Down load Speed (Mbps) Number of Tests Mean LTE Down load Speed (Mbps) Median LTE Down load Speed (Mbps) Number of Tests Mean LTE Down load Speed (Mbps) Median LTE Down load Speed (Mbps) Number of Tests Verizon Wireless 13.82 12.66 1,594 12.90 12.18 2,981 11.93 11.75 1,387 AT&T 10.94 8.98 1,245 9.95 8.23 2,543 9.06 7.56 1,298 Sprint 5.24 3.33 539 5.28 3.33 1,285 5.30 3.36 746 T-Mobile 10.66 9.54 883 10.25 9.55 1,570 9.84 9.57 688 Total 11.24 9.30 4,261 10.34 8.78 8,379 9.48 8.34 4,119 Source: The estimated speeds are based on the CalSPEED data. The top 1% of speed values were dropped, by provider and time period. Fall 2014 tests were taken between the dates of 10/1/2014 and 11/21/2014. Spring 2015 tests were taken between the dates of 5/1/2015 and 6/15/2015. Total mean and median download speeds are calculated using data for the four nationwide service providers. Federal Communications Commission DA 15-1487 138 Table VI.C.xxx CalSPEED - Estimated LTE Upload Speeds by Service Provider, California Only Service Provider Fall 2014 Fall 2014 and Spring 2015 Spring 2015 Mean LTE Upload Speed (Mbps) Median LTE Upload Speed (Mbps) Number of Tests Mean LTE Upload Speed (Mbps) Median LTE Upload Speed (Mbps) Number of Tests Mean LTE Upload Speed (Mbps) Median LTE Upload Speed (Mbps) Number of Tests Verizon Wireless 7.06 7.31 1,594 7.17 7.49 2,977 7.34 7.63 1,387 AT&T 6.44 6.63 1,245 6.23 6.36 2,540 6.06 6.14 1,298 Sprint 3.34 2.40 539 3.24 2.27 1,274 3.29 2.18 746 T-Mobile 6.49 7.23 883 6.70 7.58 1,567 7.01 7.98 688 Total 6.29 6.35 4,261 6.20 6.20 8,358 6.15 6.10 4,119 Source: The estimated speeds are based on the CalSPEED data. The top 1% of speed values were dropped, by provider and time period. Fall 2014 tests were taken between the dates of 10/1/2014 and 11/21/2014. Spring 2015 tests were taken between the dates of 5/1/2015 and 6/15/2015. Total mean and median upload speeds are calculated using data for the four nationwide service providers. Chart VI.C.i CalSPEED - Mean Downstream Throughput, 2012-Spring 2015 Source: “CALSPEED - California’s Mobile Broadband Assessment,” Spring 2015. 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Spring 2012 Fall 2012 Spring 2013 Fall 2013 Spring 2014 Fall 2014 Spring 2015 M e an D o w n st re am T h ro u gh p u t (M b /s ) Mean Downstream Throughput (Phone) AT&T Sprint T-Mobile Verizon Federal Communications Commission DA 15-1487 139 Chart VI.C.ii CalSPEED - Mean Upstream Throughput, California, 2012-Spring 2015 Source: “CalSPEED - California’s Mobile Broadband Assessment,” Spring 2015. Chart VI.C.iii CalSPEED - Mean Latency, California, 2012-Spring 2015 Source: “CalSPEED - California’s Mobile Broadband Assessment,” Spring 2015. 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Spring 2012 Fall 2012 Spring 2013 Fall 2013 Spring 2014 Fall 2014 Spring 2015 M e an U p st re am T h ro u gh p u t (M b /s ) Mean Upstream Throughput (Phone) AT&T Sprint T-Mobile Verizon 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1,000.0 Spring 2012 Fall 2012 Spring 2013 Fall 2013 Spring 2014 Fall 2014 Spring 2015 La te n cy ( m se c) Mean Latency (Phone) AT&T Sprint T-Mobile Verizon Federal Communications Commission DA 15-1487 140 APPENDIX VII: CONSUMERS AND TRENDS IN THE MOBILE WIRELESS ECOSYSTEM Table VII.C.i Percentage of U.S. Adults Living in Households with/without Wireless and Landlines (2010 - 2015) Percent of Adults in Households with: Date of interview Landline with Wireless Landline without Wireless Wireless- only Phoneless Jan–Jun 2010 62.2% 10.9% 24.9% 1.7% Jul–Dec 2010 59.4% 10.7% 27.8% 1.8% Jan–Jun 2011 58.8% 9.0% 30.2% 1.8% Jul–Dec 2011 57.3% 8.3% 32.3% 1.9% Jan–Jun 2012 56.1% 7.8% 34.0% 1.9% Jul–Dec 2012 54.4% 7.0% 36.5% 1.9% Jan–Jun 2013 52.8% 6.9% 38.0% 2.2% Jul–Dec 2013 51.5% 7.0% 39.1% 2.2% Jan–Jun 2014 47.3% 7.0% 43.1% 2.4% Jul–Dec 2014 45.8% 7.1% 44.1% 2.9% Jan–Jun 2015 43.9% 6.2% 46.7% 3.1% Note: Adults are aged 18 and over, children are under age 18. Source: CDC/NCHS National Health Interview Survey. Table VII.C.ii Percentage of U.S. Children Living in Households with/without Wireless and Landlines (2010 - 2015) Percent of Children in Households with: Date of interview Landline with Wireless Landline without Wireless Wireless- only Phoneless Jan–Jun 2010 62.8% 6.4% 29.0% 1.7% Jul–Dec 2010 59.8% 6.2% 31.8% 2.0% Jan–Jun 2011 56.7% 5.1% 36.4% 1.7% Jul–Dec 2011 54.7% 4.8% 38.1% 2.2% Jan–Jun 2012 52.7% 4.5% 40.6% 2.2% Jul–Dec 2012 49.5% 3.4% 45.0% 1.9% Jan–Jun 2013 48.3% 3.6% 45.4% 2.6% Jul–Dec 2013 46.4% 3.8% 47.1% 2.5% Jan-Jun 2014 47.1% 3.5% 52.1% 2.7% Jul–Dec 2014 39.1% 3.3% 54.1% 3.4% Jan–Jun 2015 38.3% 3.0% 55.3% 3.2% Note: Adults are aged 18 and over, children are under age 18. Source: CDC/NCHS National Health Interview Survey.