Narrowing the Digital Divide:A Better Broadband Universal ServiceProgramDaniel A. Lyons Universal service has long been an integral component of Americantelecommunications policy. As more activities move online, it becomesincreasingly important to narrow the digital divide by helping low-incomeAmericans get online and by extending broadband networks into unservedareas.Unfortunately, the Federal Communications Commission’s reforms areunlikely to help solve this problem. The Commission is repurposing an 8billion telephone subsidy program to focus instead on broadbandnetworks. But when pressed, the agency admits that it has no proof thatthe program meaningfully affected telephone adoption rates, and it offerslittle evidence that it will fare any better at boosting broadband adoption.The Federal Universal Service Fund needs revolutionary, notevolutionary, change. Rather than modifying a problematic telephone-eraprogram, Congress should adopt new initiatives designed to meet theunique challenges of the broadband era. Congress should offer targeted,direct, portable vouchers that improve the purchasing power of lowincome households and allow them to participate meaningfully intelecommunications markets. And it should create a broadbandavailability block grant program to allow state public utility commissionsto fund new networks through a reverse auction mechanism. Finally, itshould eliminate the current USF surcharge and instead fund the programdirectly, which would improve congressional oversight and minimize thefraud and abuse that has historically marred the existing program.* Copyright 2018 Daniel A. Lyons. Associate Professor of Law, Boston CollegeLaw School. Thanks to Jeff Eisenach, Brian Galle, Carolyn Gideon, Gus Hurwitz, MarkJamison, Roslyn Layton, Crystal Lyons, Randolph May, Christopher Yoo, andparticipants at the Silicon Flatirons Cable Academic Workshop for their helpfulsuggestions, as well as the Boston College Law School Fund, which helped make thisArticle possible.803

804University of California, Davis[Vol. 52:803TABLE OF CONTENTSINTRODUCTION . 805I. INTRODUCING UNIVERSAL SERVICE . 808A. The Case for Universal Service . 808B. Origins of the Universal Service Program . 8111. Universal Service Before the 1996Telecommunications Act . 8112. The 1996 Telecommunications Act . 815C. Implementing the 1996 Act: The Federal UniversalService Fund . 8171. Lifeline and Link Up . 8172. The High Cost Fund . 8193. E-rate and Rural Health Care Program . 8194. Contribution Mechanism . 820D. Transitioning the Universal Service Fund to the DigitalAge . 8211. Lifeline Reforms. 8222. High-Cost Fund and Intercarrier CompensationReform . 823II. CRITICIZING UNIVERSAL SERVICE . 826A. Lifeline . 8271. The Commission Has Not Shown the Subsidy WillBe Effective . 8272. The Lifeline Subsidy Is Incomplete . 8313. The Lifeline Subsidy Is Unnecessarily Paternalistic . 833B. Connect America Fund . 8341. The Commission Has Not Shown the Subsidy WillBe Effective . 8352. The Subsidy Is Not Technologically Neutral . 8373. The Subsidy Lacks a Long-Term Goal . 837C. E-rate . 838D. Contribution Mechanism . 839III. REFORMING UNIVERSAL SERVICE . 842A. An Activity-Based Approach to Defining BroadbandService . 842B. Low-Income Assistance . 8441. Quantifying the Drivers of the Low-IncomeBroadband Gap . 8442. Empowering Low-Income Consumers. 8453. Addressing Other Drivers of Low-IncomeNonadoption . 847

2018]Narrowing the Digital Divide8054. Shifting Away from the Federal CommunicationsCommission . 847C. Solving the Availability Gap . 849D. E-rate . 852E. Contribution Reform: Funding the Future of UniversalService . 853CONCLUSION. 854INTRODUCTIONUniversal service has long been an integral component of Americantelecommunications policy. And rightly so. Communications networksbring communities closer together and facilitate rapid exchange ofinformation among users. The purpose of universal service policies isto extend these networks to as many people as possible. And becauseof network effects, these policies benefit not only those added to thenetwork but also all other existing users, who can now reach morepeople and endpoints.1Unfortunately, the Universal Service Fund has also been one of themost criticized programs administered by the FederalCommunications Commission. And, again, rightly so. Lifeline spendsover 2 billion annually2 to bring telephone service to low-incomehouseholds, but when pressed by the Government AccountabilityOffice (“GAO”), the Commission admitted it had no proof that theprogram meaningfully increased telephone penetration rates.3 TheHigh-Cost Fund spends 4.5 billion each year4 to provide telephoneservice to rural areas, though critics argue much of these funds serveas little more than corporate subsidies for politically-connected ruralcarriers.5 More importantly, rising program costs and falling revenue1 See, e.g., Steve G. Parsons & James Bixby, Universal Service in the United States:A Focus on Mobile Communications, 62 FED. COMM. L. J. 119, 135-36 (2010).2 Lifeline Program for Low-Income Consumers, FED. COMMC’NS ogram-low-income-consumers (“The budgetwill be 2.279 billion for the calendar year beginning January 1, 2018.”) (last visitedSept. 22, 2017).3 Daniel Lyons, To Narrow the Digital Divide, the FCC Should Not Simply ExtendLifeline to Broadband, AM. ENTERPRISE INST. 1 (Mar. 30, 2016), arrow-the-Digital-Divide.pdf (last visited Sept. 22,2017).4 See FED. COMMC’NS COMM’N, UNIVERSAL SERVICE MONITORING REPORT 20 (2016)[hereinafter UNIVERSAL SERVICE MONITORING REPORT], 43025A1.pdf.5 See GOV’T ACCOUNTABILITY OFFICE, FCC HAS REFORMED THE HIGH-COSTPROGRAM, BUT OVERSIGHT AND MANAGEMENT COULD BE IMPROVED 1 (2012),

806University of California, Davis[Vol. 52:803means the universal fund surcharge has risen from 3 percent in 1998to a whopping 17.9 percent in the third quarter of 2018.6As the Internet displaces the telephone network as America’sprimary telecommunications platform, the Commission has embraceda series of reforms designed to update the universal service programfor the twenty-first century. In 2016, it enacted a comprehensivereform of the Lifeline program, allowing (and eventually requiring)recipients to repurpose their 9.25 monthly subsidy toward Internetaccess rather than traditional telephone service.7 And the Commissionbegan a slow transformation of the High-Cost Fund into a newinitiative, the Connect America Fund, the purpose of which is tosubsidize the construction of new broadband networks in unservedareas.8The Commission’s desire to promote broadband universal service isunquestionably good policy. From news to commerce to jobs andeducation, an increasing percentage of Americans’ daily activities aremoving online.9 As this transition progresses, it becomes increasinglyimportant to narrow the digital divide by helping those who cannotafford Internet access, or who live in areas where Internet access isunavailable, to get onto the network.But regrettably, the reforms it has adopted are unlikely to narrowthe digital divide. The Lifeline reforms replicate many of the problemsthat have long plagued the program: the Commission has not targetedthe subsidy to households that otherwise would not purchase Internetaccess and has offered no proof that an extra 9.25 each month wouldentice those households to buy Internet access.10 Its definition ofqualifying broadband service is inconsistent with earlier agencyrulings,11 and its desire to phase out telephone support is FED. COMMC’NS COMM’N, PROPOSED THIRD QUARTER 2018 UNIVERSAL SERVICECONTRIBUTION FACTOR 1 (2018) [hereinafter PROPOSED THIRD QUARTER 8-613A1.pdf. The factor has been as highas 19.5 % in the first quarter of 2018. FED. COMMC’NS COMM’N, PUBLIC NOTICE:PROPOSED FIRST QUARTER 2018 UNIVERSAL SERVICE CONTRIBUTION FACTOR 1 on-factor-1st-quarter-2018-195-percent.In addition to Lifeline and High-Cost support, the Universal Service Fund also spends 2.3 billion annually on E-rate, a program that provides Internet access to schools andlibraries, as well as a program to provide similar access to rural healthcare facilities.7 Lifeline & Link Up Reform & Modernization, 31 FCC Rcd. 3962, 3986 (2016).8 Connect Am. Fund, 26 FCC Rcd. 17663, 17673 (2011).9 See infra text accompanying notes 23–26.10 See infra text accompanying notes 155–84.11 See infra text accompanying notes 191–94.

2018]Narrowing the Digital Divide807unnecessarily paternalistic.12 Moreover, the Lifeline reforms do notaddress other, potentially more significant barriers to Internetadoption, such as low interest in buying household Internet access andthe high cost of computers.13 The 2016 Lifeline Reform Orderamounts to a 2.25 billion annual bet that giving a little bit of moneyto millions of low-income households will somehow solve thebroadband gap. While the Connect America Fund’s reverse auctionmechanism is a better idea, the Commission’s reforms have driven upthe Universal Service Fund’s costs without addressing the seriousstructural flaws in the program, such as lack of congressionaloversight and an unsustainable funding mechanism.14We can, and we must, do better.This Article argues that America’s approach to universal serviceneeds revolutionary, not evolutionary, change. Rather than merelytaking a flawed telephone program and extending it to broadband,Congress should design a new system to address the multifacetedchallenges posed by the digital divide.For low-income consumers, Congress should adopt acomprehensive, consumer-focused approach that addresses multipledrivers of Internet non-adoption and aims to empower low-incomefamilies to be full-fledged participants in the telecommunicationsmarketplace. This approach would encompass digital literacy outreachprograms and low-cost equipment plans as well as monthly serviceplan subsidies. The monthly plan subsidy should be data-driven,targeting only those who currently lack Internet access and providingthem with the means to secure a plan that fits the activities that weexpect recipients to be able to complete online. Consistent withPresident Obama’s ConnectALL initiative, this subsidy should bedirect and portable: recipients should receive the subsidy directly andbe able to choose how best to use this credit toward the bundle oftelecommunications services that best fit their household needs.For rural areas, Congress should adopt a block grant program inpartnership with the states. The challenges to rural broadbandconstruction are myriad and diverse, and the approach that works inflat Kansas may not be readily portable to mountainous West Virginia.State public utility commissions are in a better position than theirfederal counterparts to analyze the unique challenges faced by the121314See infra text accompanying notes 195–99.See infra text accompanying notes 185–90.See infra text accompanying notes 223–33.

University of California, Davis808[Vol. 52:803broadband industry in each state, and to design subsidy programstailored to those challenges.Finally, Congress should eliminate the opaque and unsustainablefunding mechanism for universal service. The program should beplaced on a fixed budget and subjected to congressional oversight.This would increase the incentive to deploy funds efficiently andreduce opportunities for fraud and waste that are so endemic to thecurrent statutory scheme. Congress should also consider shifting thelow-income subsidy program away from the Federal CommunicationsCommission and to another agency, such as the Department of Healthand Human Services, that better understands the drivers of povertyrelated issues.I.INTRODUCING UNIVERSAL SERVICEA. The Case for Universal ServiceThe basic tenet of universal service has been a cornerstone oftelecommunications policy for nearly a century. The 1934Communications Act charges the Federal CommunicationsCommission to “make available, so far as possible, to all the people ofthe United States . . . a rapid, efficient, Nation-wide, and world-widewire and radio communication service with adequate facilities atreasonable charges.”15 Over the years, Congress and the Commissionhave taken numerous steps to assist those who, because of geographicor socioeconomic difficulties, lack basic access to the nation’stelecommunications network.Economists often justify these policies by citing network effects. Ingeneral, the value of a network connection to a consumer generallyincreases as the number of people the consumer can reach on thenetwork increases.16 A telephone network that allows two neighbors tocall one another is somewhat useful, if they wish to speak to oneanother. But extending the network to a third neighbor enhances thevalue of that network to all three participants: the newly-addedneighbor can now use a telephone system that he could not previouslyaccess, while each of the two existing members can now call twice asmany people as before.17 A universal service policy that encourages1547 U.S.C. § 151 (2018).See Parsons & Bixby, supra note 1, at 135-36.17 See, e.g., Lifeline & Link Up Reform & Modernization, 27 FCC Rcd. 6656, 6665(2012) (“As an initial matter, all consumers, not just low-income consumers, receivevalue from the network effects of widespread voice and broadband subscribership.”).16

2018]Narrowing the Digital Divide809greater adoption rates benefits not only the direct recipients ofassistance, but all other subscribers as well.18 These network effectsdistinguish universal service from many other government entitlementprograms where the benefits to the public are less direct, and maymake it more politically palatable.19Of course, like other entitlement programs, universal service alsoconfers important benefits directly upon its recipients. In thetelephone context, perhaps the most obvious (and oft-recognized)benefit is 911 service, assuring that recipients have rapid access topotentially life-saving public safety and health care services.20 But thebenefits go far beyond simply police and fire protection. As theCommission has explained, “[t]hose consumers without affordable,quality voice services are at a disadvantage in accessing social andeconomic resources and opportunities.”21 For example, “voice serviceis particularly important for low-income consumers, who often mustjuggle multiple jobs and interviews for new employment as well askeep in contact with social service agencies.”22 Socially, telephoneservice allows a subscriber to keep in contact with friends and family,thus enhancing overall social integration.The case for a robust universal service program is even stronger inthe digital age. As more of our daily activities move online, it becomesincreasingly important to make sure that all consumers can continueto participate in society and benefit from the information revolution.These activities include:18Id.See Mark A. Lemley & David McGowan, Legal Implications of Network EconomicEffects, 86 CALIF. L. REV. 479, 546-50 (1998). Of course, these effects can beoverstated. As Lemley and McGowan explain:19While network theory holds that existing members of the telephone networkwill enjoy some benefit from the addition of new telephone users to thenetwork . . . it does not follow that this marginal private benefit exceeds orequals the marginal private cost of adding new members. Nor does it meanthat the private marginal costs and benefits will equal the costs and benefitsto society at large, even assuming, as telephony policy has decreed, thatsociety benefits as the extent of the telephone network approaches the goalof universal service.Id. at 546.20 E.g., Lifeline & Link Up Reform & Modernization, 27 FCC Rcd. at 6666 (“Voiceservice allows consumers to connect with public safety and health care resources.”).21 Id. at 6665-66.22 Id. at 6666.

University of California, Davis810[Vol. 52:803 News. Internet access lowers the cost of information,making it easier to be an informed citizen. According to arecent poll, more Americans report getting their news eachweek via laptop or computer (70 percent) than viatraditional newspapers and magazines (61 percent).23 Commerce. The ability to order goods online has yieldedsignificant consumer benefits. Former FCC ChairmanTom Wheeler notes a 2012 study showing that broadbandaccess helps a typical consumer save 8,800 each year byproviding access to bargains on goods and services.24 Jobs. Companies are increasingly leveraging the Internetto identify and recruit talent. A recent study from theCouncil of Economic Advisers shows that youngunemployed individuals who use the Internet to find jobsare re-employed 25 percent faster than those using onlytraditional methods.25 Education. Schools are increasingly integrating onlinetools in the classroom and in homework assignments. FCCCommissioner Jessica Rosenworcel has highlighted therole of Internet access for schoolchildren and the need toavoid a “Homework Gap” for those who lack access athome.26Despite these clear benefits, almost one-third of Americanhouseholds lack high-speed Internet access at home.27 By segmentingthis data, we can begin to understand the contours of the digital23 How Americans Get Their News, AM. PRESS INST. (March 17, 2014, 3:00 s/.24 Tom Wheeler, A Lifeline for Low-Income Americans, FED. COMM. COMMISSION(May 28, 2015, 1:25 PM), felinelow-income-americans.25 COUNCIL OF ECON. ADVISERS, THE DIGITAL DIVIDE AND ECONOMIC BENEFITS OFBROADBAND ACCESS 7 (2016), /files/page/files/20160308 broadband cea issue brief.pdf (citing Peter Kuhn & HaniMansour, Is Internet Job Search Still Ineffective? 18 (Inst. for the Study of Labor (Bonn),Discussion Paper No. 5955, 2011), Jessica Rosenworcel, How to Close the “Homework Gap”, MIAMI HERALD (Dec. 5,2014, 6:06 PM), 00806.html.27 JOHN B. HORRIGAN & MAEVE DUGGAN, PEW RES. CTR., HOME BROADBAND 2015, at2 (2015), http:// s/9/2015/12/Broadband-adoption-full.pdf.

2018]Narrowing the Digital Divide811divide. 67 percent of urban households, and 70 percent of suburbanhouseholds, purchase broadband access each month, but only 55percent of rural homes do so.28 The disparity is even greater whensegmented by income: 95 percent of households earning over 150,000 each year purchase broadband access, compared with onlyabout half of households earning less than 25,000.29 As the Internetdisplaces the telephone as the nation’s primary telecommunicationsnetwork, the case for modernizing our traditional universal servicemandate to fit the twenty-first century becomes increasingly strong.B. Origins of the Universal Service Program1.Universal Service Before the 1996 Telecommunications ActThe term “universal service” originated with legendary AT&Tpresident Theodore Vail, who built the Bell System and secured anationwide, state-protected telephone monopoly under the mantra of“one policy, one system, and universal service.”30 But as Professor TimWu has explained, Vail did not mean “universal as in, say, universalhealth care. Rather, it was something more akin to one universalchurch.”31 AT&T was the country’s largest telephone company,stemming from its control of Alexander Graham Bell’s originaltelephone patent.32 But it faced competition from hundreds ofindependent telephone companies that sprang up after the patentexpired in the 1890s.33 Vail argued, ultimately successfully, that thecountry was better served by a single telephone system than by amultitude of small entities.34 But in exchange for a legally protected28Id. at 8.Lifeline & Link Up Reform & Modernization, 30 FCC Rcd. 7818, 7822, 7829 &n.232 (2015); see also HORRIGAN & DUGGAN, supra note 27, at 8 (noting that 90% ofhomes earning over 100,000 per year have access, compared to only 41% of homesearning less than 20,000 per year).30 Tim Wu, The Great American Information Emperors, SLATE (Nov. 7, 2010, technology/features/2010/the greatamerican information emperors/how theodore vail built the att monopoly.html.31 Id.; see also Michael A. Janson & Christopher S. Yoo, The Wires Go to War: TheU.S. Experiment with Government Ownership of the Telephone System During World WarI, 91 TEX. L. REV. 983, 1042-44 (2013).32 See Nina Santo, Deregulation Hang Ups: Two New Laws Should JamTelecommunications “Crammers”, 30 MCGEORGE L. REV. 452, 454 (1999).33 Id. (citing MICHAEL NOLL, INTRODUCTION TO TELEPHONES & TELEPHONE SYSTEMS177 (2d ed. 1991)).34 See STUART MINOR BENJAMIN & JAMES B. SPETA, TELECOMMUNICATIONS LAW ANDPOLICY 545 (5th ed. 2015) (noting that the structure of Bell’s monopoly was accepted29

812University of California, Davis[Vol. 52:803monopoly over telephone service in much of the United States,Congress and the states imposed common carriage obligations onAT&T: the 1934 Communications Act required the company to servecustomers throughout its service area upon reasonable request, at justand reasonable rates.35During the Bell monopoly era, “universal service” referred toregulators’ efforts to manipulate the prices of various telephoneservices to serve public policy ends.36 These cross-subsidies oftenfocused on keeping basic telephone service affordable for customersthat it would not otherwise be cost-effective to service.37 For example,Bell charged commercial customers higher rates than residentialcustomers for the same service.38 By marking up commercial service,Bell could reduce the price for residential service and therefore maketelephone service more affordable for American households. Bell alsocharged similar rates to urban and rural residential customers, despitethe fact that it cost the company more to serve rural areas.39 This rateaveraging meant that urban customers cross-subsidized ruralcustomers. Finally, state and federal regulators together allowed Bell toallocate a disproportionate amount of shared network costs to longdistance service rather than local service.40 Regulators used longdistance calls, which were considered somewhat of a luxury good, tocross-subsidize local service, which was considered a necessity.Together, these policies created artificially low local service rates forresidential customers (particularly in rural areas), which helped lowand built around by federal subsidies).35 47 U.S.C. § 201 (1934); see Daniel A. Lyons, Technology Convergence andFederalism: Who Should Decide the Future of Telecommunications Regulation?, 43 U.MICH. J.L. REFORM 383, 388 (2010).36 BENJAMIN & SPETA, supra note 34, at 545.37 Daniel F. Spulber & Christopher S. Yoo, Toward a Unified Theory of Access toLocal Telephone Networks, 61 FED. COMM. L.J. 43, 48-49 (2008). As new companiesentered the market for long-distance services in the 1970s, these cross-subsidies alsobecame a tool by which Bell could stifle competition. Id. at 74.38 Id.39 BENJAMIN & SPETA, supra note 34, at 546.40 Id. The most explicit of these policies was known as the “Ozark Plan.” Enactedin 1970, the Ozark Plan “disproportionately allocated the fixed costs of the localtelephone network to long distance traffic” by requiring that for every 1 percent oftraffic consisting of long distance calls, 3.3 percent of local network costs would becovered through long distance prices. Id. at 550. So a network that carried 90 percentlocal calls and 10 percent long distance calls would allocate 67 percent of its coststoward local service and 33 percent of is costs to long distance service. Id.; see alsoPETER W. HUBER ET AL., 2004 Cumulative Supplement: Competition Policy and Antitrust,in FEDERAL TELECOMMUNICATIONS LAW 122 (2d ed. 1999).

2018]Narrowing the Digital Divide813income and rural customers who would be unable to afford service attrue cost. During the monopoly era, Bell was largely agnostic aboutthese cross-subsidies: the company was concerned primarily that itearned a reasonable rate of return overall, and was less concernedabout where the money came from.41But this carefully balanced system of cross-subsidies began tocrumble as competition displaced monopoly in portions of thetelephone industry. Savvy entrepreneurs recognized that Bell’ssupracompetitive profits in areas such as long-distance servicerepresented potential business opportunities.42 Beginning in the late1960s, MCI Communications and Sprint developed alternatives to theAT&T long distance network, marketing their private networks tocommercial clients who were being overcharged by the Bellmonopoly.43 AT&T battled these upstarts, in part because thesecompetitors’ cream-skimming threatened to undermine regulators’cross-subsidy schemes: AT&T would still be required to serve lossleading rural customers but would be unable to offset those losses inother markets that were becoming more competitive.44 Ultimately,AT&T’s efforts to combat competition drew the attention of antitrustregulators, who sued the company in 1974, alleging that AT&T wasleveraging its ongoing monopoly over local telephone exchanges todistort competition in adjacent markets in violation of the ShermanAct.45 That case settled with the 1984 Modified Final Judgment, whichbroke up the company into a competitive AT&T long-distancenetwork and seven regional, highly regulated local telephone providermonopolies known as the Regional Bell Operating Companies.46This breakup made it impossible for regulators to maintain thecross-subsidies of the monopoly era, particularly given the separationof the long-distance network from the local exchanges.47 Indeed, theModified Final Judgment’s purpose was to isolate the local telephone41See BENJAMIN & SPETA, supra note 34, at 545.Id. at 201.43 See generally Paul W. MacAvoy & Kenneth Robinson, Winning by Losing: TheAT&T Settlement and Its Impact on Telecommunications, 1 YALE J. ON REG. 1, 12 (1983).44 See BENJAMIN & SPETA, supra note 34, at 545; MacAvoy & Robinson, supra note 43.45 United States v. Am. Tel. & Tel. Co., 552 F. Supp. 131, 135 (D.D.C. 1982); seealso MacAvoy & Robinson, supra note 43, at 14.46 See Am. Tel. & Tel. Co., 552 F. Supp. at 135 (approving the settlement). TheRegional Bell Operating Companies were NYNEX, Bell Atlantic, BellSouth,AmeriTech, Southwestern Bell, US West, and Pacific Bell. Jared S. Dinkes, Note,Rethinking the Revolution: Competitive Telephony in a Voice Over Internet Protocol Era,66 OHIO ST. L.J. 833, 850 n.90 (2005).47 See BENJAMIN & SPETA, supra note 34, at 550.42

814University of California, Davis[Vol. 52:803exchanges and prevent them from entering other lines of business.48But regulators remained dedicated to keeping the cost of residentiallocal service affordable, particularly in rural areas.49 So after thebreakup, the FCC and state regulators developed a series of implicitand explicit subsidies designed to recreate the traditional crosssubsidy model within a competitive environment.50Implicit subsidies continued through a complicated access chargeregime.51 While long distance providers competed against one anotherfor residential customers, they could not complete their customers’long-distance calls without the cooperation of an originating localexchange carrier (which carried the call from the calling party’shousehold to that party’s long-distance provider) and a terminatinglocal exchange carrier (which carried the call from the long-distancenetwork to the called party’s household).52 To secure that cooperation,the long-distance company paid an access charge to the originatingand terminating local telephone exchanges.53 Because the localexchanges continued to be monopolies after the breakup, these accesscharges were regulated by tariffs: inters

In addition to Lifeline and High-Cost support, the Universal Service Fund also spends 2.3 billion annually on E-rate, a program that provides Internet access to schools and libraries, as well as a program to provide similar access to rural healthcare facilities. 7 Lifeline & Link Up R