Donate
Independent, objective, nonpartisan research

EP 707JKEP

Authors

EP 707JKEP

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3711 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(14) "EP_707JKEP.pdf" ["wpmf_size"]=> string(6) "284706" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(89508) "Public Policy Institute of California California Economic Policy is a series analyzing and discussing policy issues affecting the California economy. El l e n H a n a k , e d i tor Vo lum e 3 , N umb e r 2 n July 2 0 07 CEPCalifornia Economic Policy N early half of California households have broadband (high- speed) Internet access. Broadband is more widely available in higher-income and higher-density areas, and there are large gaps in access between the urbanized coastal regions of California and the more rural inland areas. Differences in broadband adoption rates between different racial and ethnic groups are also significant, although some of these are due to different rates of computer ownership. The technical features of broadband, including the scale economies in providing broadband infrastructure, make some regions of California more profitable to serve than others, leading to gaps in availability. Even where broadband is available, the cost of service, as well as the cost of computer hardware, results in higher rates of adoption for some than others. However, these gaps are hard to measure. This issue of California Economic Policy assesses the extent of inequalities in broadband adoption and availability in California, using an innovative method to measure its availability. All levels of government—federal, state, and local—have policies to make broadband more widely available: Policymakers hope to raise the overall level of adoption and to close the gaps between those who have access and those who do not. This report reviews the policy approaches that California and its cities are taking to raise broadband adoption and availability, includ- ing local efforts to provide municipal Wi-Fi (wireless broadband). It concludes that broadband policy in California should focus on increasing availability in rural areas and helping raise adoption rates among disadvantaged groups in urban areas. Broadband for All? Gaps in California’s Broadband Adoption and Availability By Jed Kolko S U M M A R Y 2 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A California Economic Policy Broadband for All? more available to businesses encourage job growth, increase profits, or raise productivity? Does mak- ing broadband more available to public workers lower crime rates or improve emergency response? The answers to these essential questions are largely unknown, and academic research is only beginning to approach them. nonetheless, there are several cogent argu - ments for the proposition that government should be involved in raising broadband availability and adoption and that—in the absence of government involvement—broadband use could be below a socially optimal level. on the supply side, broad - band provision involves high fixed costs, especially in rural areas; if providers were to spread the fixed cost of provision among subscribers, prices would be above marginal cost and too few people would adopt. Furthermore, broadband is most efficiently provided using publicly owned resources: Digital subscriber lines (D sl) and cable lines follow exist - ing rights-of-way, and wireless networks involve siting antennae on public property. on the demand side, there might be positive externalities in broad- band adoption, so that the benefits to society of someone adopting broadband exceed individual benefit. Also, broadband adoption encourages some online behaviors, such as looking up medical infor- mation, leading to better health outcomes—some- thing most societies consider to be a public benefit. Finally, businesses may be drawn to places where broadband is more widely available, both for bet- ter infrastructure and for a workforce that is more technologically literate, so governments consider broadband to be an economic development tool.2 The first section of this report reviews the dif- ferent aspects of the term “digital divide.” The second section explains the economics of broad- band and the technical features that could lead to geographic differences in availability. The third section outlines current broadband policy at the federal, state, and local levels. The fourth explains an alternative approach to measuring broadband availability: This involves inferring availability from adoption patterns found in a particularly rich dataset, and it overcomes important shortcomings Introduction C alifornia policymakers both locally and at the state level are undertaking numerous initiatives to raise the level of residential broadband Internet adoption in the state. Through regulation, subsidies, and direct provision, state and local governments seek to make broadband more widely available and, where available, to raise adoption rates among groups less likely to have access. However, in trying to overcome these gaps in availability and adoption, policymakers lack clear information about who in California has access to broadband and who can get it. In fact, the only comprehensive measure of availability widely used by policymakers is flawed and certainly overstates the level of broadband availability in the state. This report uses an alternative measure to assess the extent of broadband availability, adop- tion, and the digital divide within California. It seeks to answer the following questions: • Does California lead or lag the country in broadband adoption?1 • Are there inequalities in broadband availabil - ity within California? • Are there inequalities in adoption within Cali - fornia and, if so, are such inequalities more pronounced for broadband adoption or com- puter ownership? Why should overcoming a broadband digital divide, or raising the level of broadband adoption, be a policy goal? no one argues that government should boost ownership of other technologies such as DVD players and digital cameras. The difference is that broadband access (and Internet access generally) is believed to give social or economic benefits that are in the public interest. But does making broadband more available to residents improve health outcomes, lower unem- ployment, or improve job qual- ity? Does making broadband Through regulation, subsidies, and direct provision, state and local governments seek to make broadband more widely available and, where available, to raise adoption rates among groups less likely to have access. 2 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 3 California Economic Policy Broadband for All? in other measures of broadband availability tradi- tionally used in policy analysis. The fifth section presents findings about broadband availability and adoption in California. The final section draws conclusions and suggests courses of action. Gaps in Broadband Availability and Adoption A general definition of the digital divide is that it is “the gap separating those indi- viduals who have access to new forms of information technology from those who do not.”4 The digital divide encompasses disparities in avail- ability, in adoption, and in complementary skills, all of which can ultimately contribute to dispari- Text Box 1. Should We Care About Closing the Digital Divide? Some research has found positive effects of Internet use generally, although not broadband specifically, on social and economic outcomes. For example, using the Inter- net to get vehicle price information lowers costs to con- sumers by around 2 percent; furthermore, online vehicle price information eliminates the price premium that racial minorities pay of fline for new cars (see Morton, Zettel- meyer, and Silva-Risso, 2001, 2003). A separate study finds that home computer adoption is greater among people whose family or friends are more likely to use computers and, specifically, email. This finding suggests that Internet use offers a positive externality, which although not a part of the public debate about broadband policy, is the kind of justification for public spending that economists find compelling (see Goolsbee and Klenow, 2002). Other research, however, finds that the Internet does not necessarily lower consumer prices or even benefit its users. Average online book prices are no lower than in tra- ditional bookstores, and online sellers exhibit significant dispersion associated with differentiated strategies (see Clay et al., 2002). Job searchers who use the Internet do not have shorter unemployment durations than searchers who do not (see Kuhn and Skuterud, 2004). A companion study (Kolko, 2007) looks at how broadband adoption changes online behaviors—it has a positive and significant ef fect on downloading music, purchasing, visiting adult sites, and researching medica- tions and medical conditions. Adopting broadband has no statistically significant ef fect, however, on visiting job or government sites—two of the many goals that govern- ments regularly hope for when considering municipal wireless initiatives. Some studies have attempted to measure the aggre - gate economic ef fect of higher broadband adoption, focusing on the ef fect of wider broadband deployment on job grow th (especially in telecom industries respon- sible for building the infrastructure), cost savings from increased business ef ficiency, and the increase in con- sumer well-being. Estimates of the economic benefit of broadband are highly sensitive to methodology and assumptions: One study’s estimates range from $32 bil- lion to $350 billion per year in consumer surplus nation- ally, depending on assumptions about the shape of the demand cur ve (Criterion Economics, 2003).3 ties in how much benefit indi- viduals get from information technology. Furthermore, the digital divide can refer to a wide range of information technolo- gies. This report focuses on the digital divide in broadband, and this section describes how avail- ability, adoption, and complementary skills con- tribute to the broadband digital divide. This report also considers the digital divides in Internet access generally, of which broadband is one aspect, and computer ownership, which is for most people a prerequisite for adopting broadband. The first divide—that of availability—means that technologies are available for some people and not others. We discuss below why broadband Why should overcoming a broadband digital divide, or raising the level of broadband adoption, be a policy goal? could be more widely available in urban areas than in rural areas, and in richer areas than in poorer areas. The second digital divide refers to levels of adop- tion, which can also differ across groups. richer people have higher rates of broadband adoption than poorer people do. That the rich have more is not surprising, but there are two important related research questions about broadband adoption that remain to be answered. First, do race and ethnicity influence technology adoption, after differences in income and other factors are controlled for? If so, perhaps that reflects racial inequalities in technology liter- acy that policy could help over- come. Moreover, research sug- gests that disadvantaged groups can benefit disproportionately from Internet access, so targeting broadband policy to raise adop- tion rates among these groups could be especially desirable.5 s econd, how much does income matter for broadband adoption if computer owner- ship is held constant? If income affects computer ownership more than it affects broadband adop- tion, then making broadband less expensive and more widely available would have little effect on its adoption unless there were also efforts to raise computer ownership levels among lower-income people. The third digital divide concerns gaps in skills complementary to information technologies; this is often referred to as technology literacy or flu- ency. People have different levels of knowledge and comfort with technologies, and so even giving away broadband and computers would not make the benefits of information technology accessible to all if the recipients lack knowledge and familiar- ity with it. Complementary skills are not limited to technical knowledge and comfort levels with hard- ware. Two people equally familiar with technology might not reap the same benefits from an Internet connection if they have different abilities to filter California Economic Policy Broadband for All? 4 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A information— one might know which sites offer reliable and trustworthy medical advice but the other, searching for the same information, might click on the first visible sponsored link and wind up in the hands of a quack. such gaps in comple - mentary skills might or might not manifest them- selves in adoption levels. It could be that people who are less technology literate have lower demand for broadband as a consequence. Alternatively, it could be that people lacking complementary skills are no less likely to adopt broadband but benefit less from their broadband access than those with better skills. Broadband Economics T he Internet’s infrastructure consists of transmission routes, which include fiber- optic cable, coaxial cable, copper wiring, and wireless links; the infrastructure also includes connection points, where data are handed off from one route to another. An analogy with roads is use- ful: The Internet backbone is a network of high- capacity fiber-optic cables (like interstate high- ways), which connect to lower-capacity routes (like smaller highways), which in turn connect to last- mile networks (like local roads) that lead directly to residences. The Federal Communications Commission (FCC) defines broadband as an Internet service that offers speeds of at least 200 kilobits per second (kbps) in at least one direction. Most residential broadband services today offer speeds significantly faster than this, typically in the range of 1.5 mega- bits per second (Mbps: A megabit equals 1,000 kbps) to 6.0 Mbps downstream (i.e., data flow- ing from the Internet to an end-user, like a music download). upstream (i.e., data flowing from an end-user to the Internet, such as a sent email or a search request) speeds are typically slower, in the range of 384 kbps to 1.5 Mbps. By comparison, top downstream speeds over a dial-up modem are 56 kbps— only 1/100th as fast as the top of the range for broadband.6 People have different levels of knowledge and comfort with technolo- gies, and so even giving away broadband and computers would not make the benefits of information technology accessible to all. 4 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 5 California Economic Policy Broadband for All? The two primary residential broadband tech- nologies are D sl and cable. 7 Both are last-mile technologies, which means that they connect resi- dences to the larger Internet network.8 Both Dsl and cable rely on existing infrastructure to provide Internet services: D sl shares the copper wiring used for telephone service, and cable shares a hybrid of fiber and coaxial cable (HFC) used for cable televi- sion service. To offer broadband, D sl and cable providers must upgrade this existing infrastruc- ture. upgrading involves high initial fixed costs, and the technologies themselves have limitations on their deployment (described below). Because of high fixed costs and technological limitations, some cities or neighborhoods are more profitable for broadband providers to serve than others. upgrading infrastructures—and the effects on geographic availability—are different for D sl and cable.9 Dsl’s use of existing telephone cop- per wiring means that it is a dedicated service—it runs directly from the residence to the service pro- vider and is not shared with any other residences. Because D sl technology can transmit data even when the line is also in use for a telephone call, Dsl provides an always-on Internet connection. To offer D sl, telephone companies must install D sl AM (digital subscriber line access multiplexer) equipment that aggregates Internet data from the service area and forwards it to the larger high- ways of the Internet. D sl works only within three miles of a telephone company central office. Where providers offer multiple tiers of D sl service (such as 1.5, 3.0, and 6.0 Mbps), the faster tiers might be available only to residences closer to the cen- tral office. The prevalence of telephone company central offices depends on population density. large cities have multiple central offices and less- dense areas have far fewer per square mile; this alone makes D sl more widely available in higher- density areas. Alternatively, D sl architecture can consist of a copper connection from a residence to an intermediate node, called a street cabinet, which is connected with fiber to the central office. By installing street cabinets, telephone providers can offer Dsl service to areas farther than three miles from a central office.10 AT&T and Verizon are the main providers of D sl service in the united s tates, with AT&T dominant in California. Cable infrastructure consists of a head-end, which forwards local Internet traffic to the wider Internet and serves thousands of homes, and opti- cal nodes, which are connected to the head-end with fiber-optic cable and to residences with coax- ial cable. The coaxial cable is the same infrastruc- ture that delivers cable television service. To pro- vide broadband, cable television providers upgrade their networks by adding nodes and moving them closer to residences, which in effect replaces some of the coaxial cable in the network with fiber. Cable companies also have to install equipment (analogous to the telephone D slAMs) that route and switch digital data, and they install amplifi- ers that improve the upstream data transmission.11 u nlike D sl, cable infrastructure is shared: The coaxial cable connects residences to the optical node in a loop, so that the bandwidth any resi- dence receives depends in part on the number of residences sharing the node. Thus, the local fixed costs of upgrading an area’s infrastructure to make cable broadband available can involve (1) moving optical nodes closer to residences, (2) building new optical nodes, (3) upgrading the upstream path, and (4) installing equipment at the head-end.12 Comcast and Time Warner are the main providers of cable broadband service in California. These fi xed costs mean that cable and D sl are more profitable in areas where the costs can be spread over more subscrib- ers. A reas that are higher den- sity or higher income or both tend to be more profitable. In a higher-income neighbor- hood, more residents are likely to adopt broadband, so provid- ers seeking to make broadband available in the most profitable areas first would choose higher- income areas; this effect is mag- nified for cable providers, since upgrading their networks offers new potential revenue from both Because of high fixed costs and technological limitations, some cities or neighborhoods are more profitable for broadband providers to serve than others. broadband and digital television service. High- density neighborhoods are not only more likely to be within three miles of a telephone central of fice for Dsl service, they also can be served at lower installation costs per subscriber.13 For all of these reasons, broadband availabil- ity should differ geographically according to aver- age income and density. There might also be geo- graphic differences in availability that are unique to a particular provider: Most areas in the united s tates are served by a dominant telephone pro - vider and a dominant cable provider, and each can make different strategic decisions about when to introduce broadband service to their regions. The age and physical condition of existing telephone or cable infrastructure can also affect the cost of introducing broadband. In addition to these two wireline technologies, two wireless broadband technologies are begin- ning to be used.14 o ne is satellite, which although available nearly everywhere in the united states, offers a slower speed and lower reliability for a higher monthly price than either cable or D sl. 15 In prac- tice, satellite broadband appeals to consumers only where D sl and cable are unavailable and so does not actually compete with them.16 The other wireless broadband technology is Wi-Fi (the technical term is 802.11x), which offers high speed within a very short distance of a base station. Wi-Fi is commonly used in conjunc- tion with D sl or cable to make broadband access available wirelessly within a home, office, café, or public space. The cost of setting up base stations and antennas to provide Wi-Fi coverage is much lower than upgrading or building wireline infra- structures such as D sl, cable, or fiber-to-the-home (FTTH). In a handful of locations, Wi-Fi offers citywide public access to the Internet and actually competes with cable and D sl. This new wave of municipal Wi-Fi initiatives is discussed below as an example of broadband policy. California Economic Policy Broadband for All? 6 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Current Broadband Policies— Federal, State, Local F ederal, state, and local governments all play a role in shaping the availability and adop- tion of broadband. The federal government, through Congress, the executive, and the supreme Court, makes the most important regulatory deci- sions, whereas state and local governments play a larger role in subsidizing and in some cases directly providing broadband services. At the federal level, the FCC regulates tele- communications. The 1996 Telecommunications Act, the 2005 supreme Court decision in National Cable & Telecommunications Association et al. v. Brand X Internet Services et al., and related FCC rulings have created the regulatory framework that exists today. under this framework, telephone, wireless, television, and Internet providers are able to “compete in any market against any other” (FCC web site), but broadband providers are not required to give competitors wholesale access to their infrastructures so that competitors can resell services to consumers.17 The FCC also shapes broadband policy by placing conditions on merg- ers between broadband providers and by allocating wireless spectrum.18 Finally, the FCC administers the universal service requirement, which guaran- tees that even the most remote areas have telephone service.19 s tates and local governments are still left with important elements of broadband regulation, one being control of the physical development of infra- structure. Broadband networks typically follow public rights-of-way such as roads or rail tracks, and deploying broadband infrastructure costs providers less when it occurs in tandem with pub- lic works projects or when done simultaneously by multiple broadband providers. In California, improving rights-of-way access for broadband deployment is the most prominent element of Governor Arnold schwarzenegger’s 2006 Execu - tive order on “Expanding Broadband Access and u sage in California,” which created a California Federal, state, and local governments all play a role in shaping the availability and adoption of broadband. California Economic Policy Broadband for All? 6 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 7 Text Box 2. Will New Internet Access Technologies Overcome the Digital Divide? The economics of DSL and cable result in greater geographic differences in availability than those of dial-up access do. Will the next generation of technologies have the same effect or will they widen or narrow the digital divide? The most promising next-wave broadband technol - ogies are FT TH and WiMax.20 FT TH extends fiber-optic cable—which already connects the Internet all the way to the telephone company central of fices and the cable pro- viders’ optical nodes—the rest of the way to homes. Fiber would, in ef fect, replace the copper wiring and coaxial cable in use today. Because the capacity of fiber is far greater than that of either copper or coaxial cable, users would access much higher speeds downstream and upstream—potentially into the gigabit-per-second (gbps) range, hundreds of times faster than today’s fastest cable or DSL ser vices.21 Recent policy reforms at the state and federal levels have reduced some of the regulator y challenges to deliv- ering cable television and these give telephone companies a stronger incentive to build FT TH networks.22 Telephone companies are interested in FT TH as much for its abil- ity to deliver television ser vice as to deliver high-speed Internet ser vice; this is so that they can compete fully with cable providers, who can of fer television, Internet, and telephony (using voice-over-IP [Internet protocol]) over their existing networks.23 However, current adoption in California is minimal. SureWest, a regional telecom ser vice provider, has over 20,000 FT TH subscribers in the Sacramento area.24 AT&T, the dominant local phone provider, has not announced plans for any major FT TH deployment. Fiber costs more than DSL or cable, and it requires replacement of the existing connections to customers’ homes, so its roll-out proceeds neighborhood by neighbor- hood. Multiunit dwellings, dense areas, and new develop- ments cost less to wire with fiber than other areas, so there is a strong possibility of a future, persistent digital divide in FTTH availability; at least as important is that only some telecom companies are considering FTTH. In fact, a study prepared for San Francisco’s evaluation of the feasibility of a municipally built, owned, and operated fiber-optic network argues that the city is already on the losing side of the FTTH digital divide because Verizon, the company deploying most large FTTH projects in the country, is not the dominant telephone provider in San Francisco.25 The other promising next-wave technology is WiMax, a wireless technology. WiMax of fers Internet connectiv- ity over a range of one to 30 miles from a transmission tower; this compares to the hundreds of feet that are Wi- Fi’s limit. The downstream and upstream bandwidth of WiMax depends on the number of simultaneous users, but speeds could rival those of DSL and cable. WiMax is not yet being used for broad-based Internet access. Theo- retically, the fixed costs of WiMax deployment should be much lower per subscriber than costs for wireline tech- nologies, because the infrastructure consists of widely spaced antennas, not extensive wiring. The wide range of WiMax signals could also bring high-speed ser vice to harder-to-reach rural areas. Next-wave technologies are no guarantee of over - coming today’s digital divide. In fact, the high fixed costs of FT TH make it likely that some areas will receive ser- vice long af ter others do. Fur thermore, with the devel- opment of faster access technologies, expectations about adequate ser vice ratchet upward. This is not only because the digital divide refers to relative dif ferences, not absolute levels. It is also because online applications are designed for users’ current bandwidth; as typical resi- dential bandwidth increases, online applications incorpo- rate more bandwidth-hungr y content (such as video and interactivity), and access technologies that were once adequate cease to be so. California Economic Policy Broadband for All? 8 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A silicon Valley. Major initiatives are under negotia- tion for san Francisco, for the greater sacramento region, and for a silicon Valley–wide network; los Angeles has also announced a citywide initiative.31 Most of these call for low-cost or free wireless access, provided by partners such as Google, Earth- l ink, and MetroFi, supported by subscriptions or advertising, with little financial investment by the public sector. of the 58 initiatives, all but four are in densely populated areas of the Bay Area, south - ern California, or greater sacramento. Text box 3 provides more detail on the justifications and chal- lenges of municipal wireless. California’s broadband policy can be summed up as follows: First, subsidies focus more on insti- tutional access to broadband than residential access to broadband. second, municipal Wi-Fi initiatives are widespread and are concentrated in the densely populated parts of the state. Third, policy is geared toward broadband access, not computer ownership. Measuring Broadband T rying to measure the extent of the digital divide is challenging. Publicly available household surveys on broadband adop- tion, Internet access, and computer ownership are inadequate for studying recent trends in Califor- nia. For example, the federal Current Population Survey last included technology questions in 2003, and there are no plans to do so again. The Pew Internet & American life Project surveys house - holds about technology adoption annually or more frequently, but with only 4,000 respondents nationally, the sample is too small to draw con- clusions about California.32 A proprietary survey, the Technographics Benchmark conducted by For- rester, a technology research and consulting firm, is used for this analysis. Forrester annually surveys 60,000 –100,000 households about their technol- ogy adoption and behaviors.33 Measuring the divide in broadband availabil- ity is more challenging than measuring broadband adoption. Broadband providers treat service avail- Broadband Task Force composed of public and private stakeholders to coordinate efforts to raise broadband adoption and identify ways to fund new technology investments.26 s tates and localities also play a large role through subsidizing and directly providing broad- band. California’s primary broadband subsidy program is the Teleconnect Fund, which pays half the cost of Internet access for qualified schools, libraries, community organizations, and other nonprofits. Funded from a statewide fee on tele- phone service, the fund’s 2006 –2007 fiscal year budget is $22 million.27 In addition, the Califor- nia Public utilities Commission (CP uC) recently created the California Emerging Technology Fund, an independent nonprofit foundation to be funded with $60 million over five years from AT&T and Verizon as conditions of their respective mergers with sBC and MCI. The fund’s mission is “achiev - ing ubiquitous access to broadband and advanced services in California, particularly in underserved communities through the use of existing and emerg- ing technologies,” although specific strategies have not yet been selected.28 Whereas state governments have focused on subsidies to encourage adoption, more and more localities are attempting to provide broadband directly, both by themselves and in partnership with private companies. In the late 1990s, a few localities across the country built fiber-optic net- works.29 These early projects often involved public owner- ship of networks and were in direct response to the perceived lack of service provision by the phone and cable companies.30 In the past couple of years, many localities have turned to Wi-Fi as a wireless standard and a way to bring broadband service at low or no cost to a wide area. In Cal- ifornia, 58 localities have Wi-Fi initiatives under way. service is operational in Anaheim, the san Diego County tribal nations, and several cities in Whereas state govern- ments have focused on subsidies to encourage adoption, more and more localities are attempting to provide broadband directly, both by them- selves and in partnership with private companies. California Economic Policy Broadband for All? 8 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 9 Text Box 3. Local Wi-Fi Initiatives As of early 2007, numerous localities are developing wireless broadband networks using Wi-Fi technology to ser ve entire regions. In contrast to earlier ef for ts, most municipal Wi-Fi plans and deployments today call for at least par tial ownership and operation by the private sec- tor. So phone and cable companies, af ter fighting earlier attempts at direct public provision, are instead par tner- ing with local governments.34 The best-known are in Phil- adelphia, where a wireless network run by Ear thLink is operational, and in San Francisco, which is still negotiat- ing with Ear thLink and Google; across the countr y there are hundreds of others under way.35 Many arguments over municipal wireless rest on technical issues specific to Wi-Fi, which was not designed to provide city wide coverage but to bring it to a build- ing, park, or other small area. Wi-Fi transmits signals over relatively shor t distances (up to 30 meters indoors, 450 meters outdoors) and is the technology behind pub- lic “hotspots” and home networks. Municipal networks extend the capabilities of Wi-Fi by using multiple trans- ceiver sites that collectively cover a large area. The first technical concern about Wi-Fi is that it is unclear how far such a Wi-Fi signal can reach indoors, and users might need to install range-extending equipment. Second, new wireless standards such as WiMax that can transmit sig- nals much far ther could make Wi-Fi obsolete. Third, a city wide Wi-Fi signal could interfere with existing Wi-Fi hotspots.36 There is also controversy on social and econom - ic grounds. In San Francisco, the par tnership calls for Ear thLink to provide paid access and for Google to pro- vide free, adver tising-suppor ted access; Google’s model, which uses tracking cookies that customize adver tising, has raised privacy concerns.37 In some cities, the busi- ness model itself is being debated: Should wireless be adver tising-suppor ted (and free), or subscription-based, or a hybrid?38 Some are even questioning the public-pri- vate par tnership model: as of May 2007, San Francisco’s Board of Super visors has delayed finalizing the agree- ment with Ear thLink and Google to assess a city-owned and -operated alternative. Despite these controversies, there is broad consensus that the benefits of municipal Wi-Fi include (1) narrowing the digital divide and (2) facilitating online activities that are socially desirable or economically productive. Large cities such as Philadelphia and San Francisco focus most on the digital divide among residents and on bringing free or low-cost access to ever yone. Philadelphia, for in- stance, negotiated with Ear thLink to provide broadband at a lower cost to lower-income residents.39 Large cities also want to encourage socially desirable online behav- iors related to health care, education, and employment.4 0 Promoters of Silicon Valley’s request for proposal for a wireless network, however, emphasize the digital divide among businesses. Some are out of reach of both DSL and cable providers, they argue, and desirable online activities include those that improve business develop- ment, government ser vices, and public safety.41 Still unknown is how much demand there is among residents, businesses, and visitors for municipal Wi-Fi. Municipal Wi-Fi competes with existing fixed-wire con- nections in homes and workplaces; it also competes with data ser vices of fered by mobile phone providers, which of fer Internet connectivity to mobile phones and to spe- cially equipped laptops. Taipei has one of the world’s most extensive wireless networks, reaching 90 percent of the city’s 2.6 million people, but six months af ter its launch in early 2006, only 40,000 residents had subscribed.42 California Economic Policy Broadband for All? 10 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A ability maps as sensitive competitive information, and regulators have not required that provid- ers make these maps public. Although individual consumers can check availability at an individual address on most providers’ web sites, anecdotal evidence suggests that this information is often inaccurate or incomplete.43 The only widely used data on broadband avail- ability come from FCC surveys of broadband pro- viders. The FCC publishes a count of the number of providers with at least one subscriber in each zip code nationally.44 This method overstates the level of availability because broadband may be available in only some parts of a zip code. using this measure could also under- state geographic differences in broadband availability if provid- ers make broadband available only in richer or denser portions of a zip code. recent FCC data suggest implausibly high levels of broadband availability, ren- dering the data misleading for assessing the digital divide. The FCC reports that in December 2005, 99.8 percent of the u.s . population lived in a zip code where broadband was available (see Table 1). In the most sparsely populated tenth of zip codes, 96.2 percent of the population lived in zip codes where broadband was available, and in the poorest tenth of zip codes, 99.4 percent of the population lived in zip codes where broadband was available. nevertheless, these FCC data are the basis for most policy studies and academic work about broadband availability. In its most recent report on broadband deployment in the united states in 2004, the FCC concludes that broadband is nearly univer- sally available in urban areas and “significant prog- ress is being made towards ubiquitous availability of advanced services in rural areas.”45 The California Public utilities Commission (2006) also relies on the FCC’s zip code data for its maps of broadband availability, which show at least one broadband pro- vider offering service in every zip code in Califor- nia.46 s everal academic studies rely on the FCC data as well, although one notes that “the FCC count of high speed line providers within a zip code may seri- ously misrepresent competitive options available to the totality of residents within that zip code … [but] there is no practical alternative to using the FCC data in assessing broadband availability.”47 r ather than rely on FCC data, this analy - sis infers broadband availability by examining Table 1. U.S. Broadband Availability, FCC Measure Population in Zip Codes with At Least One High-Speed Subscribera (%) Population of Lowest-Density Zip Codes (bot tom decile) in Zip Codes with At Least One High-Speed Subscriber (%) Population of Lowest-Income Zip Codes (bot tom-decile) in Zip Codes with At Least One High-Speed Subscriber (%) December 2 0 0 09 6.44 9.991.5 December 2 0 01 97.8b67.995.1 December 2 0 0 29 9.18 0.997.5 December 2 0 0 39 9.58 8.99 8.6 December 2 0 0 49 9.691.89 9.0 December 2 0 059 9.89 6.29 9.4 source: Federal Communications Commission (2006), Tables 18 and 19. aBased on FCC data and the author’s calculations.bCompared to an online Forrester survey in autumn 2001, in which 64 percent of online respondents reported broadband availability where they live. Rather than rely on FCC data, this analysis infers broadband availability by examining the relation- ship between location and broadband adoption at the individual level, controlling for numerous individual characteristics. growth in broadband adoption is not yet leveling off. Throughout this period, the level of broadband adoption in California has been above the national rate. Most recently, 47 percent of California house- holds had broadband, eight points higher than the overall national level. There are several possible explanations for this. o ne is that California residents have more favor - able demographics for broadband adoption—that is, they are richer, more educated, and so on—than Americans generally, and these demographic factors influence broadband adoption. A second possibility is that broadband is more widely available in California because the density of population and income levels make it more prof- itable to offer broadband here than elsewhere in the country. A third possibility is that another California-specific factor—such as state policies or decisions by AT&T, the dominant local tele- phone provider in California— contributed to faster deployment and therefore higher adoption. These three possibilities are not mutually exclusive. California Economic Policy Broadband for All? 10 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 11 Looking just at the 68 percent of U.S. house- holds with Internet access (broadband or dial-up) at home, more than half have broadband. the relationship between location and broadband adoption at the individual level, controlling for numerous individual characteristics.48 This tech- nique, in essence, compares people who are identi- cal in observable demographics, such as income, age, education, size of household, age of children, and so on, but who live in different types of zip codes—for example, a middle-income urban zip code and a middle-income rural zip code.49 Because their individual demographics are the same, they are assumed to have similar underlying demand for broadband, and any difference in whether they actually have broadband is interpreted as a differ- ence in the availability of broadband in the differ- ent types of zip codes.50 The key location charac- teristics are zip code log median household income and zip code log population density.51 Measuring the digital divide in complementary skills presents further difficulties. self-perceptions of skills are harder to elicit and interpret, so research on complementary skills often relies on direct obser- vation and usability testing, which is expensive to conduct on a large scale. Furthermore, as technol- ogy changes, particular skills rise or fall in impor- tance, making measurement and interpretation even more difficult.52 n or do any of the main data sources on access or availability, including Forrester, solicit detailed information on complementary skills, so this aspect of the digital divide is beyond the scope of this report. Broadband Availability and Adoption in California O ur first research question is whether Cali- fornia leads or lags the nation in broad- band adoption. nationally, residential broadband adoption has grown rapidly, from 7 percent of households at the end of 2000 to 39 per- cent in 2005 (see Figure 1).53 l ooking just at the 68 percent of u.s . households with Internet access (broadband or dial-up) at home, more than half have broadband. During 2005, broadband adop- tion grew from 28 percent to 39 percent, so the Figure 1. Broadband Adoption in the United States and California Source: Forrester Research and author’s calculationsNote: Data based on survey conducted by mail in English only. Percentage 200420032002200120002005 45 40 35 30 25 20 15 10 5 50 0 California United States 47% 39% 10%7% 15% 18% 28% 10% 17% 23% 24% California Economic Policy Broadband for All? 12 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A The results of this analysis suggest that about half of the gap between broadband adoption in California and that in the rest of the nation is due to California’s favorable individual demographics. To assess the importance of each, regression anal- ysis can identify the separate relationships of individ- ual demographics and location characteristics such as neighborhood income and density on broadband adoption. The results of this analysis suggest that about half of the gap between broadband adoption in California and that in the rest of the nation is due to California’s favorable individual demograph- ics.54 l ocation characteristics that affect providers’ decisions to make broadband available—neighbor- hood income and population den- sity—account for the other half of the gap.55 Because individual demographics and location char- acteristics account for the entire difference between broadband adoption in California and that in the rest of the country, we infer that other factors unique to California did not collectively change the level of broadband adoption.56 The second research question is whether there are inequalities in broadband availability within California. As noted above, this report infers avail- ability from the relationship between location characteristics and broadband adoption. Another regression analysis reveals that broadband adop- tion in California is significantly higher in higher- income and denser zip codes, even after controlling for individual demographics.57 That is, people of the same age, race, income, and so on are more likely to adopt broadband if they live in a richer or denser area.58 The effect of location on avail- ability means that broadband adoption differs con- siderably across regions within California, some of which are much richer and denser than others. Broadband adoption ranges from under 30 percent in the north of the state and in the sierras to just over 50 percent in the Bay Area and greater los Angeles (see Table 2).59 The differences in Inter- net access (broadband and dial-up combined) and computer ownership across regions are smaller: Internet access ranges from 61 percent to 76 per- cent and computer ownership ranges only between 73 percent and 83 percent. More than half the dif- ferences in broadband adoption between regions remain even after controlling for individual demo- graphics, suggesting that inequalities in broadband availability across regions account for much of the differences in broadband adoption. In contrast, the differences between regions in overall Internet access and computer ownership are due to differ- ences in individual demographic characteristics across those regions.60 The third research goal is to understand the inequalities in broadband adoption in California apart from availability. Although it is obvious (as we confirm below) that broadband adoption rates are higher among richer households, two impor- tant questions about income and broadband adop- tion whose answers are not obvious are (1) do race and ethnicity affect broadband adoption, holding income constant, and (2) does income affect broad- band adoption because income also affects com- puter ownership, which is essentially a prerequisite for broadband adoption?61 Broadband adoption, Internet adoption, and computer ownership all vary by income in Cali- fornia. Among households with incomes over $100,000, 68 percent have broadband, compared to 49 percent of households with incomes between $50,000 and $75,000 and compared to 24 percent of households with incomes under $25,000 (see Table 3). The gaps for Internet adoption (broad- band and dial-up combined) and computer owner- ship across income groups are also considerable. looking across racial and ethnic groups, it is important to keep in mind that Forrester conducted this survey in English only. These data therefore represent people with very high English proficiency and exclude a significant share of California’s Hispanic population.62 Disparities in broadband adoption across racial and ethnic groups are less wide than across income groups but they are still apparent, with 46 percent of both non-Hispanic whites and English-proficient Hispanics having broadband, compared with 63 percent of Asian Americans and only 36 percent of African Ameri- cans.63 s ince income among non-Hispanic whites is English and spanish, broadband adoption is 50 percent among Hispanics who speak English as much or more than spanish at home and responded to the sur- vey in English. Broadband adop- tion is only 20 percent among Hispanics who speak spanish more than English at home or responded to the survey in span - ish. Broadband adoption among all California Hispanics is 29 percent, so looking only at His- California Economic Policy Broadband for All? 12 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 13 higher than income among Hispanics, and income and broadband adoption are strongly related, why do both groups have the same level of broadband adoption in California? relative to non-Hispanic whites, Hispanics on average have larger house- holds and are younger, both of which are positively correlated with broadband adoption. other research shows that Hispanics in Cali - fornia who prefer speaking spanish at home have much lower broadband adoption than English- speaking Hispanics. According to the March 2007 PPIC statewide survey, conducted by telephone in Table 2. Broadband, Internet, and Computer Ownership, by California Region, 2005 Broadband (%) Online (broadband or dial-up) (%)Computer (%)Number Nor thern California2 96 37 7161 Nor thern Sacramento Valley2 8697916 3 Greater Sacramento4 4768 3518 San Francisco Bay Area51748 01, 3 3 5 Nor thern San Joaquin Valley3 66774241 Southern San Joaquin Valley3 56 4733 8 7 Central Sierra2161765 4 Central Coast4 8768 0176 Greater Los Angeles5276811,9 4 9 Inland Empire4 573826 6 0 San Diego Border4 8727872 2 s ources: Forrester r esearch and the author’s calculations. n otes: Counties in each region are listed in the web-only appendix at www.ppic.org /content /other/707J KEP_web_ appendix.pdf. Data are based on a survey conducted by mail in English only. Table 3. Broadband, Internet, and Computer Ownership in California, by Income, 2005 Household Income ($ 1,000s)Broadband (%) Online (broadband or dial-up) (%)Computer (%)Number < 2 5244 85 81,167 2 5 – 4 94 0707 71,573 5 0 – 694 9788 61,14 6 70 – 9 9598 7911, 32 0 10 0 +6 88 99 31, 3 82 sources: Forrester r esearch and the author’s calculations. n ote: Data are based on a survey conducted by mail in English only. Broadband adoption ranges from under 30 percent in the north of the state and in the Sierras to just over 50 percent in the Bay Area and greater Los Angeles. California Economic Policy Broadband for All? 14 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A panics with high English proficiency overstates the level of broadband adoption among Hispanics.64 Holding other individual demographics con- stant, there is a statistically significant difference for broadband adoption among Hispanics and African Americans relative to whites. English- preferring Hispanics are 6 percent less likely to have broadband at home than non-Hispanic whites; the difference for African Americans is 10 percent.65 r elative to California’s overall broad- band adoption of 47 percent, these gaps are large. Moreover, for Hispanics this is the gap between adoption among non-Hispanic whites and Hispan- ics with high English proficiency. The racial gaps in broadband adoption, however, cannot be explained by provid- ers’ failing to offer service in minority neighborhoods because there was no statistically signifi- cant effect of the racial composi- tion of a zip code on broadband adoption. The other important fac- tor about the digital divide in broadband adoption is the role of computer ownership. A home computer is for the most part a prerequisite for broadband adoption.66 If the relationship between income and broadband adoption reflects lower computer own- ership by poorer households, then making broad- band more widely available at low cost might have little effect on broadband adoption. looking across income groups, computer ownership ranges from 58 percent among households with income under $25,000 to 93 percent among households with income $100,000 and above. regression analysis helps assess how much this gap in computer own- ership contributes to the digital divide in broad- band adoption. Controlling for other demograph- ics, raising log income by one standard deviation increases the likelihood of computer ownership by 7 percent.67 Then, looking only at computer own- ers and again controlling for other demographics, raising income by the same amount increases the likelihood of broadband adoption by 10 percent.68 Both relationships are statistically significant: The effect of income on computer ownership means that there is a digital divide in computer owner- ship, but the effect of income on broadband among computer owners means that, even among com- puter owners, there is a digital divide in broadband adoption. Conclusions and Recommendations W ithin California, there are digital divides in both broadband availability and broadband adoption. Broadband is more widely available in higher-density residen- tial areas and urban regions than in lower-density areas and rural regions; the level of neighborhood income also affects availability. These inequali- ties arise, in part, from the technical features and economic realities of broadband provision. The digital divide in adoption encompasses not only broadband but computer ownership as well, and income and race and ethnicity both affect broad- band adoption in California. our findings have important implications for broadband policy. If closing gaps in broadband availability is a policy goal, raising availability in rural areas should be the top priority. The Califor- nia Emerging Technology Fund should focus first on deployment in rural areas, and the Broadband Task Force could help identify clearly the barriers to providers’ offering service in rural areas. Eas- ing the access to rights-of-way might facilitate rural broadband deployment, but the economics of broadband make rural areas costly to serve even in the absence of any regulatory or permitting fac- tors, so the state could consider direct subsidies to providers serving rural areas. Although broadband availability lags in lower-income areas as well, den- sity has a greater effect on availability; furthermore, there is no evidence that broadband availability is lower in neighborhoods with higher concentrations of Hispanics and African Americans. Holding other individual demographics constant . . . English-preferring Hispanics are 6 percent less likely to have broadband at home than non-Hispanic whites; the difference for African Americans is 10 percent. California Economic Policy Broadband for All? 14 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 15 To close gaps in broadband adoption, the focus should be broadened to include racial and ethnic and income groups with lower adoption rates, including those living in neighborhoods where broadband service is already available. Improving technology literacy may also be essential. Munici- pal Wi-Fi initiatives will help meet these goals so long as they provide broadband at lower cost than Dsl and cable providers do. recent initiatives sug - gest that they will: Google’s proposal to provide free, ad-supported service in san Francisco, and the prevailing rate of about $20 per month that Earth link charges for its Wi-Fi service, are two examples. Municipalities need not resort to pub- lic ownership or operation of Wi-Fi networks to achieve these public benefits. Public involvement should be reserved for the investments that the pri- vate sector is less likely to make, such as helping raise technology literacy and improving computer access among disadvantaged residents, both of which are among the “digital inclusion” goals of san Francisco’s Wi-Fi initiative. 69 Ironically, because the Wi-Fi networks that are operational or under consideration are over- whelmingly in the urban parts of the state, they are likely to widen rather than close the gap in avail- ability between urban and rural areas. Policymak- ers, therefore, should think less in terms of closing the gap between urban and rural availability, lest they hold back urban initiatives, and more in terms of raising the absolute level of rural availability. n ew technologies, such as FTTH, are also likely to benefit urban areas more than rural areas, so the urban-rural gap is likely to persist even if policy efforts successfully raise the level of rural broad- band availability. Two broader questions remain to be researched and answered. First, is it more desirable from a cost-benefit perspective to raise the degree of broadband adoption for those with less access than it is to overcome other inequalities that Californians face? on one hand, the economic and social benefits of broadband are unknown and are largely taken as an article of faith; on the other, the costs to the public sector of raising broadband adoption can be low if the private sector bears the cost, as is the case with most municipal Wi-Fi initiatives. sec - ond, will the development of future Internet access technolo- gies such as WiMax or FTTH mitigate or aggravate today’s inequalities in tech- nology access? Both of these questions should be considered— even if they cannot be answered— as California’s state and local governments weigh various strategies for raising broadband adoption and closing the digital divide. v If closing gaps in broadband availability is a policy goal, raising availability in rural areas should be the top priority. To close gaps in broad- band adoption, the focus should be broadened to include racial and ethnic and income groups with lower adoption rates. California Economic Policy Broadband for All? 16 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Notes 1 The California Public utilities Commission (2005) cites California’s #14 ranking among states (according to the 2003 Tech net survey) as an important indicator of wheth-er the state is “maintaining its lead in broadband usage.” In the California Public utilities Code, the main broad-band policy objectives are assuring “continued affordabili-ty and widespread availability” of broadband and promot-ing “economic growth, job creation, and substantial social benefits.” It is hard to see how the “substantial social ben-efits” depend on California’s broadband adoption relative to that of other states, rather than the absolute level, but it is plausible that the relative ranking could affect economic growth and job creation if businesses consider broadband availability and adoption a factor in deciding where to open, expand, or move their operations. 2 o nly limited research has been completed on the effects of broadband. An overview of the academic literature on the effects of broadband and Internet usage is presented in Text Box 1. 3 California Public utilities Commission (2005) summarizes several studies on broadband and economic development. 4 Gunkel (2003). 5 see Morton, Zettelmeyer, and s ilva-risso (2003). 6 Dial-up Internet access, in contrast to cable and D sl, is nearly ubiquitous in the united states and never had a long period of geographic disparities in availability. Mak-ing dial-up available does not involve a high fixed cost: Dial-up Internet service providers (I sPs) need to establish a point-of-presence (P oP) in an area by making a local phone number available as a dial-in number, and although this involves adding some switching equipment in the tele-phone company central office, the infrastructure invest-ment is minimal. Downes and Greenstein (2002) docu-ment that dial-up Internet service spread quickly to even the most rural counties: In 1997, 99 percent of the u.s . population lived in counties with at least one I sP, and 92 percent lived in counties with seven or more I sPs. There are scale economies in having a P oP, but the costs to upgrade infrastructure for broadband service are much larger (Greenstein and Prince, 2006). 7 D sl refers to a family of technologies, technically de- scribed as xD sl. residential D sl is most often AD sl (asymmetric D sl), and higher-capacity variations include s D sl and V D sl. 8 Many smaller businesses subscribe to D sl as well. larger businesses rely on higher-bandwidth connections such as T1 and T3 lines. Cable broadband is almost exclusively a residential service. 9 The description of Internet infrastructure and the process for upgrading networks relies on Federal Communications Commission (2000) and Corning (2001, 2005). 10 The website www.dslreports.com/prequal/distance offers estimates of distance between a residential address and the nearest central office and maps of central office locations. 11 Traditional cable television is a one-way communica-tion: Video is broadcast from the cable head-end to the residence. Internet service, as well as digital cable service, is two-way communication, which requires a different “upstream” infrastructure. 12 Cable providers might have to make additional infra-structure investments as broadband adoption rises, con-gestion increases, and bandwidth speeds fall. 13 High residential density is one reason why Korea, Hong Kong, and Japan have higher broadband adoption and higher speeds of service available than the united states does. 14 In addition to satellite and Wi-Fi, there is also WiMax, which is discussed in Text Box 2. 15 satellite also has a higher upfront equipment cost. Hughes net basic service costs $59.99 per month for speeds up to 700 kbps plus a $400 equipment and installation fee. see the website go.gethughesnet.com. 16 s ee “With a Dish, Broadband Goes r ural” (2006). 17 The laws and rulings that make up the federal regula-tion of broadband have involved the sometimes competing goals of (1) maximizing competition among broadband providers, (2) not reducing the incentives that broadband providers have to invest in infrastructure and new applica- tions, and (3) treating D sl and cable broadband services evenly despite the fact that telephone companies as “com-mon carriers” have traditionally been subject to much stricter federal regulation than cable companies have. 18 In 2005, the FCC required that AT&T offer “naked D sl ,” under which consumers can subscribe to AT&T’s D sl service without subscribing to AT&T’s local telephone service, as a condition of its merger with sBC. Then, in late 2006, the FCC required specific pricing for naked D sl as a condition of its merger with Bell south. see “AT&T to o ffer ‘ naked D sl’ for Far l ess Than Before” (2007). 19 A short explanation of universal service is available at the FCC’s website, www.fcc.gov/cgb/consumerfacts/ universalservice.html. universal service applies only to telephone, not Internet, service. However, because D sl uses the telephone network, the cost of providing D sl in remote areas is lower than if the telephone network did not cover those areas. In contrast, cable companies face no universal service requirement. As a result, Dsl adoption is higher than cable broadband adoption in very low-density rural areas. 20 Fiber-to-the-home is one technology in a class of FTTx technologies: They differ in whether the fiber network extends all the way to the individual home, or to the curb (FTTC), or to a multiunit building (FTTB). Fiber networks include those that have dedicated fiber running between the California Economic Policy Broadband for All? 16 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 17 customer and the service provider (point-to-point networks) or have shared fiber (as in a “passive optical network”). 21 Another emerging wireline technology is Broadband over Powerline (BP l), which uses the electric power network and home electrical wiring and outlets to carry Internet data. unlike FTTH, BP l speeds are similar to those of D sl and cable (see the website www.fcc.gov/cgb/consumerfacts/highspeedinternet.html). 22 In 2006, California passed a law creating statewide video franchises, which allow television providers to apply for a single statewide franchise rather than multiple local franchises as cable companies have traditionally had to do. Also in 2006, the FCC put time limits on the local franchise negotiation process and on franchise fees; it also curtailed unreasonable “build-out” requirements, which specify how quickly television providers are required to offer service to the entire franchise area. see the C nET.com article at http://news.com.com / FCC+adopts+relief+for+ telecom+companies+planning+T V+offerings/2100 -1036 _3- 6145184.html. 23 Verizon’s television-over-IP service is called Fi os. AT&T delivers its television service, u-Verse, by extending its fi-ber network closer to, but not all the way to, residences and then delivering content over the last portion of the network using a higher-capacity version of D sl. see the C n ET.com article at http://news.com.com /AT T+to+ramp+ up+IPT Vs+expansion /2100 -1037_ 3- 6153354.html. 24 s ee sureWest’s 2006 Q3 financial statement, at www.su r w.com /med ia _ relat ions /press /releases /ea r n i ngs /pd f /Q3_06financials_p6.pdf. 25 n one of san Francisco’s broadband providers plan to deploy FTTH to san Francisco beyond limited trials, and these providers’ future plans are “not comparable” to FTTH (Columbia Telecommunications Corporation, 2007, p. 2). Verizon, in contrast, plans to spend $20 billion over several years to make fiber-to-the-home high-speed services avail-able to 60 –70 percent of customers in its current service area, which includes some areas in California, according to “Verizon’s Fios services Build Momentum” (2006). 26 The Executive order designates the Business, Transpor-tation, and Housing Agency as the lead agency for most items, highlighting the importance of infrastructure devel- opment to the goal of broadband access. on many telecom policy issues, including rate regulation, the Public utilities Commission has responsibility. 27 California Public utilities Commission (2005, p. 62) has more details of the program and also of the similar federal e- rate program. 28 s ee the fund’s website at www.cetfund.org. 29 In Tacoma, Washington, the municipal power utility, City light, built, owns, and operates a fiber-optic network that delivers Internet and television. see “Cities Deliver Broadband for l ess” (2003). 30 Gillett, lehr, and osorio (2004) review many of these local broadband initiatives. 31 A full list of municipal wireless initiatives in the united s tates is available at www.muniwireless.org. status of ini-tiatives comes from the January 2007 summary. 32 The most recent Pew report on broadband adoption is based on a survey of 4,000 adults over the period February–April 2006. see www.pewinternet.org /pdfs/ PIP_ Broadband_trends2006.pdf. 33 Forrester’s annual Technographics Benchmark survey is conducted by mail, in English only; the samples are selected from national market research panels to be rep- resentative of u.s . households demographically and are weighted to correct for differences in response rates. For- rester has used the T ns (formerly nFo ) market research panel since 2001 and used nPD’s panel in earlier years. Forrester collects data in the 48 contiguous states and the District of Columbia but not in Alaska or Hawaii. some respondents participate in Forrester’s survey in multiple years: Kolko (2007) includes some longitudinal analysis from these data. Brown and Goolsbee (2002), Goolsbee (2000), Goolsbee and Klenow (2002), and Prince (2003) use Forrester’s data as well. The author was employed at Forrester from 2000 to 2005. 34 “Companies That Fought Cities on Wi-Fi, now rush to Join In” (2006). 35 A comprehensive list is available at www.muniwireless.com. 36 “Wi-Pie in the sky” (2006) reviews the technological and managerial challenges of citywide Wi-Fi networks. 37 “s ome Worries as san Francisco Goes Wireless” (2006). 38 Google provides free Wi-Fi in Mountain View, Cali-fornia, and MetroFi offers both free and paid Wi-Fi in Cupertino, sunnyvale, and santa Clara, California. see “ s.F. Picks Google Wi-Fi Team” (2006), and “Google Gives City Free Wi-Fi” (2006). 39 In Philadelphia, Earth link charges $21.95 per month generally and $9.95 per month for low-income households. 40 s ee, for instance, the literature from Wireless Philadelphia, available at www.phila.gov/wireless/briefing.html; from s an Francisco TechConnect, available at www.sfgov.org / site/techconnect_tf_index.asp; and from an interview with Houston’s chief information officer at www.govtech.net/digitalcommunities/story.php?id=98722. 41 “Public, Private Collaboration To Design silicon Val- ley Wide Wireless network” (2006). silicon Valley, unlike most large cities, has a lower share of low-income residents and more geographically dispersed businesses that aren’t well served by D sl or cable providers. California Economic Policy Broadband for All? 18 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 42 s ee “What If They Built an urban Wireless network and Hardly Anyone used It?” (2006). The article cites the competition from Wi-Fi connections, mobile phone data services, and fixed-wire broadband as a major reason for the lack of interest in paying for Taipei’s municipal Wi-Fi service. 43 A personal anecdote: over a two-day period in August 2006 of checking the AT&T website and calling several customer service departments to sign up for new D sl ser- vice, the author was told by different people that (1) D sl was unavailable at the address, (2) only a slower D sl ser -vice (up to 1.5 Mbps) was available at the address, (3) all speeds of Dsl service were available at the address, and (4) the address was not a valid address. 44 These counts include both providers who own the broad- band pipes (such as the dominant telephone and cable companies) and resellers. Although residential custom-ers account for most of the dominant telephone and cable companies’ subscribers, some resellers serve primarily business customers. 45 FCC (2004), p. 38. The FCC uses “advanced services” as a synonym for broadband. Two of the five FCC com-missioners dissented from the upbeat conclusions of the report, and both noted the inadequacy of using the one- subscriber-in-a-zip-code measure of availability. still, the main concern of both dissenters was the lower level of broadband adoption in the united states relative to that in Korea, Japan, and other countries, not the inequality of broadband availability within the united states. Further-more, not only do global broadband leaders have a higher percentage of households with broadband, average broad-band speeds are also much higher. 46 The CP uC emphasizes the inequality in the number of providers in zip codes, rather than using the data as evidence of ubiquitous access to broadband, and notes that “this rep-resentation does not depict the availability of broadband for every resident in each identified zip code area.” 47 Flamm (2006). Grubesic (2006), and Prieger (2003) also use FCC broadband data. 48 This section of the report draws heavily on Kolko (2007). Details on the methodology for inferring broadband avail-ability can be found there. 49 u nlike the CP s or Pew, Forrester includes respondents’ zip codes. 50 This approach raises two methodological concerns. First, location characteristics could be capturing unobserved indi- vidual characteristics. second, location itself could affect the demand for broadband if, for instance, people with less access to doctors or stores have greater demand for online health advice or online shopping. Kolko (2007) finds that location characteristics have a much smaller or insignifi-cant effect on technologies other than broadband and that location characteristics strongly influence the choice of cable versus D sl, which is much more plausibly due to availability differences than to location-driven differences in underlying demand for one broadband technology ver- sus the other. still, it is important to keep in mind that this approach measures availability as the relationship between location characteristics and adoption, which might also be capturing some effects in addition to availability. Further-more, this measure does not capture reasons other than zip code density and zip code income for variation in availabil-ity, such as zip-code-level topographical differences that might make some zip codes more expensive to serve. 51 This approach implicitly captures geographic differences in broadband prices. If there is more competition among broadband providers in higher-density areas, this could result in lower prices and therefore higher adoption. The notion of “availability” used here is not just a binary mea-sure but rather a continuous measure that captures dimen-sions of availability such as price and quality of service. 52 DiMaggio et al. (2004) review the academic literature on the digital divide in complementary skills and provide more detail on the challenges researchers face in studying it. 53 Forrester’s measure of broadband refers to “high-speed Internet access” and mentions D sl and cable modem as examples. In other questions in Forrester’s surveys that ask about specific technologies, broadband includes D sl, cable, fixed wireless, and satellite. 54 That is, the coefficient on the California dummy vari-able in a probit regression where broadband adoption is the dependent variable falls by about half when individual demographics are included. 55 That is, the coefficient on the California dummy variable in a probit regression where broadband adoption is the dependent variable is not significantly different from zero when individual demographics and zip code characteris-tics (log median income and log density) are included. 56 Because there could be many factors unique to Califor-nia that affect broadband adoption, one cannot conclude that they are irrelevant individually. For instance, if, hypo-thetically, AT&T had been unusually aggressive in rolling out broadband in California, but state policy made deploy-ment more difficult than in other states, these two effects could cancel each other out. Thus, we emphasize that fac-tors unique to California did not collectively have an effect on broadband adoption. 57 The web-only appendix www.ppic.org /content/other/707JKEP_web_appendix.pdf presents regression results for California and also for the united states. Higher-income and higher-density zip codes have higher broadband adop-tion, controlling for individual characteristics, both for California and the united states. The effect of income is smaller for California than for the united states, although statistically significant in both cases. The effect of density is similar for California and the u nited states. California Economic Policy Broadband for All? 18 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 19 58 These significant and positive relationships do not hold when Internet access (broadband or dial-up) is the depen-dent variable, or when computer ownership is the depen-dent variable. This is further evidence that the relation-ship between zip code characteristics and adoption is due to availability, since the availability of dial-up access and computer ownership do not vary geographically. relative to the overall broadband adoption level of 47 percent, liv-ing in a zip code with a log density one standard deviation (1.72) above the mean raises the likelihood of broadband adoption by six percentage points—an amount not only statistically significant but large in magnitude as well. 59 Table 2 shows the nine regions defined by the Economic s trategy Panel, but we have split the san Joaquin Valley into northern and southern regions, and we have split the south - ern California region into Greater los Angeles and Inland Empire. The distribution of Forrester’s respondents across these regions is similar to the population distribution. 60 That is, an F-test does not reject the hypothesis that the regional dummies all equal zero in a regression of Inter-net adoption (and, separately, computer ownership) on individual characteristics and regional dummies. This test does reject the hypothesis of insignificant regional dum-mies when broadband adoption is the dependent variable. 61 Although this section focuses on race, ethnicity, and in-come, it is also a goal of California policy to consider the digital divide between people with disabilities and people without. Forrester’s surveys do not ask about disability status. 62 The Forrester Technographics survey is long, detailed, and conducted by mail, so respondents need considerable English proficiency to complete it. Furthermore, both His-panics and African Americans are underrepresented in Forrester’s survey relative to their shares in the population reported by the Census. However, the income distribu-tions for these groups in Forrester’s survey are quite close to those in the March 2005 Current Population survey, so Forrester’s sample appears to be a reasonable enough representation for us to report results. 63 some non-Hispanic whites, African Americans, and Asian Americans also lack English proficiency, but the English proficiency requirement should affect the repre-sentativeness of the Hispanic sample most. 64 Because Hispanics preferring to speak spanish are a considerable share of California’s population, the overall level of broadband adoption in California rises from 55 percent among all Californians to 65 percent when these Hispanics are excluded, suggesting that the Forrester figure for English-speaking Californians would fall by up to 10 percentage points if non-English-speakers were also sur-veyed. These figures are higher than the Forrester figures for overall adoption in part because the PPIC survey was conducted 15 months after the Forrester survey. Also, the PPIC data, unlike Forrester data, show a gap in adoption between Hispanics preferring to speak English and whites, in part because a higher level of English proficiency would be required to complete Forrester’s mail survey in English than to participate in PPIC’s telephone survey in English. An earlier survey, the october 2003 Current Population Survey, reports that broadband adoption is 18 percent for California Hispanics who are not spanish-only versus only 2 percent for spanish-only Hispanics in California. Data from PPIC and the Current Population Survey are weighted using standard demographics to represent the population more accurately, but the figures reported here do not further adjust for demographics using a regression framework. 65 The regression underlying this finding also controls for zip code income and density. 66 Why might a household have broadband without a com- puter? some videogame consoles, digital video record-ers, and smartphones can connect directly to the Internet using cable or D sl. Households could also have broad-band access for use with an employer-provided computer. In California, 80 percent of households have a computer. Among computer-owning households, 58 percent have broadband at home. Among households with broadband, 98 percent own a computer. 67 r aising log income by one standard deviation at the mean of log income is equivalent to raising income from $48,000 to $114,000. The standard deviation of log income is .875. 68 These results are from (1) a probit regression of computer ownership on individual characteristics and (2) a probit regression of broadband adoption on individual and zip code characteristics, conditional on computer ownership. 69 Another strategy for increasing computer ownership is manufacturing and distributing extremely low-cost com- puters, which is the goal of the nonprofit one laptop Per Child foundation, www.laptop.org. Its focus is raising computer use among children in developing countries. California Economic Policy Broadband for All? 20 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Federal Communications Commission, Availability of Advanced Telecommunications Capacity in the United States: Second Report to Congress, Washington, D.C., 2000. Federal Communications Commission, Availability of Advanced Telecommunications Capacity in the United States: Fourth Report to Congress, Washington, D.C., 2004. Federal Communications Commission, High Speed Ser-vices for Internet Access: Status as of December 31, 20 05, Washington, D.C., July 2006. Flamm, Kenneth, “Diagnosing the Disconnected: Where and Why Is Broadband Access unavailable in the u.s .?” unpublished draft, 2006. Gillett, sharon, William lehr, and Carlos osorio, “ local Government Broadband Initiatives,” Telecommunications Policy, Vol. 28, 2004, pp. 537–558. “Google Gives City Free Wi-Fi,” San Francisco Chronicle, August 16, 2006. Goolsbee, Austan, “In a World Without Borders: The Impact of Taxes on Internet Commerce,” Quarterly Jour-nal of Economics , Vol. 115, no. 2 , 2000, pp. 561–576. Goolsbee, Austan, and Peter Klenow, “Evidence on learn - ing and network Externalities in the Diffusion of Home Computers,” Journal of Law and Economics, Vol. 45, 2002 , pp. 317–343. Greenstein, shane, and Jeff Prince, “The Diffusion of the Internet and the Geography of the Digital Divide in the u nited states,” nBE r Working Paper #12182 , Cambridge, Massachusetts, 2006. Grubesic, Tony, “A spatial Taxonomy of Broadband r egions in the united states,” Information Economics and Policy, Vol. 18, 2006, pp. 423 – 448. Gunkel, David, “ second Thoughts: Toward a Critique of the Digital Divide,” New Media & Society, Vol. 5, 2003, pp. 499–522. Kolko, Jed, “Why should Governments support Broadband Adoption?” Public Policy Institute of California, Working Paper #2007.01, s an Francisco, California, 2007. Kuhn, Peter, and Mikal skuterud, “Internet Job search and unemployment Durations,” American Economic Review , Vol. 94, n o. 1, March 2004, pp. 218 –232. Morton, Fiona scott, Florian Zettelmeyer, and Jorge silva- r isso, “Internet Car retailing,” Journal of Industrial Economics , Vol. 49, n o. 4, 2001, pp. 501–519. Bibliography “AT&T to offer ‘ naked D sl’ for Far less Than Before,” San Francisco Chronicle, January 16, 2007. Brown, Jeffery, and Austan Goolsbee, “Does the Internet Make Markets More Competitive? Evidence from the life Insurance Industry,” Journal of Political Economy, Vol. 110, n o. 3, 2002 , pp. 481–507. California Public utilities Commission, “Broadband Deployment in California,” san Francisco, California, 2005. California Public utilities Commission, “Broadband Deployment in California,” san Francisco, California, 2006. “Cities Deliver Broadband for less,” Wired News, March 7, 2003, available at www.wired.com /news/ business/0,1367,57927,00.html. Clarke, richard, Kevin Hassett, Zoya Ivanova, and lau -rence Kotlikoff, “Assessing the Gains from Telecom Com- petition,” nBE r Working Paper #10482 , Cambridge, Massachusetts, May 2004. Clay, Karen, ramayya Krishnan, Eric Wolff, and Danny Fernandes, “ retail strategies on the Web: Price and n on-Price Competition in the online Book Industry,” Journal of Industrial Economics , Vol. 50, no. 3, 2002 , pp. 351–367. Columbia Telecommunications Corporation, “Fiber o ptics for Government and Public Broadband: A Feasibil- ity study,” prepared for the City and County of san Fran-cisco, Columbia, Maryland, 2007. “Companies That Fought Cities on Wi-Fi, now rush to Join In,” Wall Street Journal, March 20, 2006. Corning, Inc.,“Bridging the last Mile: Access network Wireline Architectures,” White Paper W P 6300, Corning, new y ork, 2001. Corning, Inc., “Broadband Technology overview,” White Paper W P 6321, Corning, n ew york, 2005. Criterion Economics, “The Effects of ubiquitous Broad - band Adoption on Investment, Jobs, and the u.s . Econ-omy,” Washington, D.C., 2003. DiMaggio, Paul, et al., “From unequal Access to Differen - tiated use,” in Kathryn neckerman, ed., Social Inequality, r ussell s age Foundation, n ew york, n ew york, 2004. Downes, Tom, and shane Greenstein, “ understanding Why universal service obligations May Be unnecessary: The Private Development of local Internet Access Mar -kets,” unpublished draft, 2006. California Economic Policy Broadband for All? 20 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 21 “s.F. Picks Google Wi-Fi Team,” San Francisco Chronicle, April 6, 2006. “some Worries as san Francisco Goes Wireless,” New York Times, April 10, 2006. “Verizon’s Fios services Build Momentun,” C nET, August 1, 2006, available at http://news.com.com / Verizons+Fios+services+build+momentum /2100 -1034_ 3- 6101038.html. “What If They Built an urban Wireless network and Hardly Anyone used It?” New York Times, June 26, 2006. “Wi-Pie in the s ky,” The Economist, March 9, 2006. “With a Dish, Broadband Goes rural,” New York Times, n ovember 14, 2006. Morton, Fiona scott, Florian Zettelmeyer, and Jorge silva- r isso, “Consumer Information and Discrimination: Does the Internet Affect the Pricing of new Cars to Women and Minorities?” Quantitative Marketing and Economics, Vol. 1, 2003, pp. 65–92. Prieger, James, “The supply side of the Digital Divide: Is There Equal Availability in the Broadband Internet Access Market?” Economic Inquiry , Vol. 41, no. 2 , April 2003, pp. 346 –363. Prince, Jeff, “Measuring the Digital Divide: structural Estimation of Demand for Personal Computers,” unpub-lished manuscript, 2003. “Public, Private Collaboration To Design silicon Valley Wide Wireless network,” Joint Venture silicon Valley press release, January 26, s an Jose, California, 2006. About the Author Jed Kolko is a research fellow at the Public Policy Institute of California. Contributors Davin Reed provided research assistance for this edition of California Economic Policy. I would like to thank Richard Greene for editorial assistance and Dean Bonner and Mark Baldas-sare for PPIC sur vey data. I am grateful to Forrester Research for permission to use the Techno-graphics Benchmark data. Anne Neville (State of California Business, Transpor tation, and Housing Agency), Shane Greenstein (Nor thwestern Universit y), Dan Mazmanian (Universit y of Southern California and PPIC), Jaime Calleja Alderete (PPIC), Ellen Hanak (PPIC), Deborah Reed (PPIC), and Mark Baldassare (PPIC) provided detailed and thought ful comments on draf t versions of this repor t. Brian Moura ( Wireless Silicon Valley Initiative) provided background insights on municipal wireless initiatives. Par ticipants at the Telecommunications Policy Research Conference and at PPIC seminars of fered helpful suggestions. Responsibilit y for any errors is solely mine. Walter B. Hewlet tDirectorCenter for Computer Assisted Research in the Humanities Ki Suh ParkDesign and Managing Par tnerGruen Associates Constance L . Rice Co-DirectorThe Advancement Project Raymond L . WatsonVice Chairman of the Board EmeritusThe Ir vine Company Carol WhitesidePresidentGreat Valley Center Board of Directors Thomas C. Sut ton, ChairRetired Chairman and Chief E xecutive Of ficer Pacific Life Insurance Company Mark BaldassarePresident and Chief E xecutive Of ficerPublic Policy Institute of California Linda GriegoPresident and Chief E xecutive Of ficerGriego Enterprises, Inc. Edward K. HamiltonChairmanHamilton, Rabinovit z & Alschuler, Inc. Gary K. Har tFormer State Senator and Secretar y of EducationState of California The Public Policy Institute of California is a private, nonprofit research organization established in 19 9 4 with an endowment from William R. Hewlet t . The Institute conducts independent, objective, nonpartisan research on the economic, social, and political issues af fecting Californians. The Institute’s goal is to raise public awareness of these issues and give elected representatives and other public of ficials in California a more informed basis for developing policies and programs. Copyright © 20 07 by Public Policy Institute of CaliforniaAll rights reser vedSan Francisco, CA Shor t sections of tex t , not to exceed three paragraphs, may be quoted without writ ten permission provided that full at tribution is given to the source and the above copyright notice is included. PPIC does not take or suppor t positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, suppor t , or oppose any political par ties or candidates for public of fice. Research publications reflect the views of the authors and do not necessarily reflect the views of the staf f, of ficers, or Board of Directors of the Public Policy Institute of California. PUBLIC POLICY INS TIT U TE OF CALIFORNIA5 0 0 Washington Street , Suite 8 0 0 San Francisco, California 9 4111Telephone: (415) 2 91- 4 4 0 0 Fax: (415) 2 91- 4 4 01 w w w.ppic.org ISSN #1553-8737 California Economic Policy Broadband for All? PuBl IC Pol ICy Ins TITuTE o F CAlIF orn IA 500 Washington s treet, suite 800 s an Francisco, California 94111NON-PROFIT ORG.U.S. POSTAGE PAIDBRISBANE, CAPERMIT #83 In This Issue of CEP Who has broadband? Who can get it? are available free of charge on PPIC’s website www.ppic.org R E C E N T I S S U E S O F California Economic Policy Pay-or-Play Health Insurance Mandates: Lessons from California Lawns and Water Demand in California Trade with Mexico and California Jobs Are Businesses Fleeing the State? Interstate Business Relocation and Employment Change in California A Decade of Living Wages: What Have We Learned?" } ["___content":protected]=> string(106) "

EP 707JKEP

" ["_permalink":protected]=> string(118) "https://www.ppic.org/publication/broadband-for-all-gaps-in-californias-broadband-adoption-and-availability/ep_707jkep/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8608) ["ID"]=> int(8608) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:39:06" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3845) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(10) "EP 707JKEP" ["post_type"]=> string(10) "attachment" ["slug"]=> string(10) "ep_707jkep" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(14) "EP_707JKEP.pdf" ["wpmf_size"]=> string(6) "284706" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(89508) "Public Policy Institute of California California Economic Policy is a series analyzing and discussing policy issues affecting the California economy. El l e n H a n a k , e d i tor Vo lum e 3 , N umb e r 2 n July 2 0 07 CEPCalifornia Economic Policy N early half of California households have broadband (high- speed) Internet access. Broadband is more widely available in higher-income and higher-density areas, and there are large gaps in access between the urbanized coastal regions of California and the more rural inland areas. Differences in broadband adoption rates between different racial and ethnic groups are also significant, although some of these are due to different rates of computer ownership. The technical features of broadband, including the scale economies in providing broadband infrastructure, make some regions of California more profitable to serve than others, leading to gaps in availability. Even where broadband is available, the cost of service, as well as the cost of computer hardware, results in higher rates of adoption for some than others. However, these gaps are hard to measure. This issue of California Economic Policy assesses the extent of inequalities in broadband adoption and availability in California, using an innovative method to measure its availability. All levels of government—federal, state, and local—have policies to make broadband more widely available: Policymakers hope to raise the overall level of adoption and to close the gaps between those who have access and those who do not. This report reviews the policy approaches that California and its cities are taking to raise broadband adoption and availability, includ- ing local efforts to provide municipal Wi-Fi (wireless broadband). It concludes that broadband policy in California should focus on increasing availability in rural areas and helping raise adoption rates among disadvantaged groups in urban areas. Broadband for All? Gaps in California’s Broadband Adoption and Availability By Jed Kolko S U M M A R Y 2 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A California Economic Policy Broadband for All? more available to businesses encourage job growth, increase profits, or raise productivity? Does mak- ing broadband more available to public workers lower crime rates or improve emergency response? The answers to these essential questions are largely unknown, and academic research is only beginning to approach them. nonetheless, there are several cogent argu - ments for the proposition that government should be involved in raising broadband availability and adoption and that—in the absence of government involvement—broadband use could be below a socially optimal level. on the supply side, broad - band provision involves high fixed costs, especially in rural areas; if providers were to spread the fixed cost of provision among subscribers, prices would be above marginal cost and too few people would adopt. Furthermore, broadband is most efficiently provided using publicly owned resources: Digital subscriber lines (D sl) and cable lines follow exist - ing rights-of-way, and wireless networks involve siting antennae on public property. on the demand side, there might be positive externalities in broad- band adoption, so that the benefits to society of someone adopting broadband exceed individual benefit. Also, broadband adoption encourages some online behaviors, such as looking up medical infor- mation, leading to better health outcomes—some- thing most societies consider to be a public benefit. Finally, businesses may be drawn to places where broadband is more widely available, both for bet- ter infrastructure and for a workforce that is more technologically literate, so governments consider broadband to be an economic development tool.2 The first section of this report reviews the dif- ferent aspects of the term “digital divide.” The second section explains the economics of broad- band and the technical features that could lead to geographic differences in availability. The third section outlines current broadband policy at the federal, state, and local levels. The fourth explains an alternative approach to measuring broadband availability: This involves inferring availability from adoption patterns found in a particularly rich dataset, and it overcomes important shortcomings Introduction C alifornia policymakers both locally and at the state level are undertaking numerous initiatives to raise the level of residential broadband Internet adoption in the state. Through regulation, subsidies, and direct provision, state and local governments seek to make broadband more widely available and, where available, to raise adoption rates among groups less likely to have access. However, in trying to overcome these gaps in availability and adoption, policymakers lack clear information about who in California has access to broadband and who can get it. In fact, the only comprehensive measure of availability widely used by policymakers is flawed and certainly overstates the level of broadband availability in the state. This report uses an alternative measure to assess the extent of broadband availability, adop- tion, and the digital divide within California. It seeks to answer the following questions: • Does California lead or lag the country in broadband adoption?1 • Are there inequalities in broadband availabil - ity within California? • Are there inequalities in adoption within Cali - fornia and, if so, are such inequalities more pronounced for broadband adoption or com- puter ownership? Why should overcoming a broadband digital divide, or raising the level of broadband adoption, be a policy goal? no one argues that government should boost ownership of other technologies such as DVD players and digital cameras. The difference is that broadband access (and Internet access generally) is believed to give social or economic benefits that are in the public interest. But does making broadband more available to residents improve health outcomes, lower unem- ployment, or improve job qual- ity? Does making broadband Through regulation, subsidies, and direct provision, state and local governments seek to make broadband more widely available and, where available, to raise adoption rates among groups less likely to have access. 2 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 3 California Economic Policy Broadband for All? in other measures of broadband availability tradi- tionally used in policy analysis. The fifth section presents findings about broadband availability and adoption in California. The final section draws conclusions and suggests courses of action. Gaps in Broadband Availability and Adoption A general definition of the digital divide is that it is “the gap separating those indi- viduals who have access to new forms of information technology from those who do not.”4 The digital divide encompasses disparities in avail- ability, in adoption, and in complementary skills, all of which can ultimately contribute to dispari- Text Box 1. Should We Care About Closing the Digital Divide? Some research has found positive effects of Internet use generally, although not broadband specifically, on social and economic outcomes. For example, using the Inter- net to get vehicle price information lowers costs to con- sumers by around 2 percent; furthermore, online vehicle price information eliminates the price premium that racial minorities pay of fline for new cars (see Morton, Zettel- meyer, and Silva-Risso, 2001, 2003). A separate study finds that home computer adoption is greater among people whose family or friends are more likely to use computers and, specifically, email. This finding suggests that Internet use offers a positive externality, which although not a part of the public debate about broadband policy, is the kind of justification for public spending that economists find compelling (see Goolsbee and Klenow, 2002). Other research, however, finds that the Internet does not necessarily lower consumer prices or even benefit its users. Average online book prices are no lower than in tra- ditional bookstores, and online sellers exhibit significant dispersion associated with differentiated strategies (see Clay et al., 2002). Job searchers who use the Internet do not have shorter unemployment durations than searchers who do not (see Kuhn and Skuterud, 2004). A companion study (Kolko, 2007) looks at how broadband adoption changes online behaviors—it has a positive and significant ef fect on downloading music, purchasing, visiting adult sites, and researching medica- tions and medical conditions. Adopting broadband has no statistically significant ef fect, however, on visiting job or government sites—two of the many goals that govern- ments regularly hope for when considering municipal wireless initiatives. Some studies have attempted to measure the aggre - gate economic ef fect of higher broadband adoption, focusing on the ef fect of wider broadband deployment on job grow th (especially in telecom industries respon- sible for building the infrastructure), cost savings from increased business ef ficiency, and the increase in con- sumer well-being. Estimates of the economic benefit of broadband are highly sensitive to methodology and assumptions: One study’s estimates range from $32 bil- lion to $350 billion per year in consumer surplus nation- ally, depending on assumptions about the shape of the demand cur ve (Criterion Economics, 2003).3 ties in how much benefit indi- viduals get from information technology. Furthermore, the digital divide can refer to a wide range of information technolo- gies. This report focuses on the digital divide in broadband, and this section describes how avail- ability, adoption, and complementary skills con- tribute to the broadband digital divide. This report also considers the digital divides in Internet access generally, of which broadband is one aspect, and computer ownership, which is for most people a prerequisite for adopting broadband. The first divide—that of availability—means that technologies are available for some people and not others. We discuss below why broadband Why should overcoming a broadband digital divide, or raising the level of broadband adoption, be a policy goal? could be more widely available in urban areas than in rural areas, and in richer areas than in poorer areas. The second digital divide refers to levels of adop- tion, which can also differ across groups. richer people have higher rates of broadband adoption than poorer people do. That the rich have more is not surprising, but there are two important related research questions about broadband adoption that remain to be answered. First, do race and ethnicity influence technology adoption, after differences in income and other factors are controlled for? If so, perhaps that reflects racial inequalities in technology liter- acy that policy could help over- come. Moreover, research sug- gests that disadvantaged groups can benefit disproportionately from Internet access, so targeting broadband policy to raise adop- tion rates among these groups could be especially desirable.5 s econd, how much does income matter for broadband adoption if computer owner- ship is held constant? If income affects computer ownership more than it affects broadband adop- tion, then making broadband less expensive and more widely available would have little effect on its adoption unless there were also efforts to raise computer ownership levels among lower-income people. The third digital divide concerns gaps in skills complementary to information technologies; this is often referred to as technology literacy or flu- ency. People have different levels of knowledge and comfort with technologies, and so even giving away broadband and computers would not make the benefits of information technology accessible to all if the recipients lack knowledge and familiar- ity with it. Complementary skills are not limited to technical knowledge and comfort levels with hard- ware. Two people equally familiar with technology might not reap the same benefits from an Internet connection if they have different abilities to filter California Economic Policy Broadband for All? 4 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A information— one might know which sites offer reliable and trustworthy medical advice but the other, searching for the same information, might click on the first visible sponsored link and wind up in the hands of a quack. such gaps in comple - mentary skills might or might not manifest them- selves in adoption levels. It could be that people who are less technology literate have lower demand for broadband as a consequence. Alternatively, it could be that people lacking complementary skills are no less likely to adopt broadband but benefit less from their broadband access than those with better skills. Broadband Economics T he Internet’s infrastructure consists of transmission routes, which include fiber- optic cable, coaxial cable, copper wiring, and wireless links; the infrastructure also includes connection points, where data are handed off from one route to another. An analogy with roads is use- ful: The Internet backbone is a network of high- capacity fiber-optic cables (like interstate high- ways), which connect to lower-capacity routes (like smaller highways), which in turn connect to last- mile networks (like local roads) that lead directly to residences. The Federal Communications Commission (FCC) defines broadband as an Internet service that offers speeds of at least 200 kilobits per second (kbps) in at least one direction. Most residential broadband services today offer speeds significantly faster than this, typically in the range of 1.5 mega- bits per second (Mbps: A megabit equals 1,000 kbps) to 6.0 Mbps downstream (i.e., data flow- ing from the Internet to an end-user, like a music download). upstream (i.e., data flowing from an end-user to the Internet, such as a sent email or a search request) speeds are typically slower, in the range of 384 kbps to 1.5 Mbps. By comparison, top downstream speeds over a dial-up modem are 56 kbps— only 1/100th as fast as the top of the range for broadband.6 People have different levels of knowledge and comfort with technolo- gies, and so even giving away broadband and computers would not make the benefits of information technology accessible to all. 4 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 5 California Economic Policy Broadband for All? The two primary residential broadband tech- nologies are D sl and cable. 7 Both are last-mile technologies, which means that they connect resi- dences to the larger Internet network.8 Both Dsl and cable rely on existing infrastructure to provide Internet services: D sl shares the copper wiring used for telephone service, and cable shares a hybrid of fiber and coaxial cable (HFC) used for cable televi- sion service. To offer broadband, D sl and cable providers must upgrade this existing infrastruc- ture. upgrading involves high initial fixed costs, and the technologies themselves have limitations on their deployment (described below). Because of high fixed costs and technological limitations, some cities or neighborhoods are more profitable for broadband providers to serve than others. upgrading infrastructures—and the effects on geographic availability—are different for D sl and cable.9 Dsl’s use of existing telephone cop- per wiring means that it is a dedicated service—it runs directly from the residence to the service pro- vider and is not shared with any other residences. Because D sl technology can transmit data even when the line is also in use for a telephone call, Dsl provides an always-on Internet connection. To offer D sl, telephone companies must install D sl AM (digital subscriber line access multiplexer) equipment that aggregates Internet data from the service area and forwards it to the larger high- ways of the Internet. D sl works only within three miles of a telephone company central office. Where providers offer multiple tiers of D sl service (such as 1.5, 3.0, and 6.0 Mbps), the faster tiers might be available only to residences closer to the cen- tral office. The prevalence of telephone company central offices depends on population density. large cities have multiple central offices and less- dense areas have far fewer per square mile; this alone makes D sl more widely available in higher- density areas. Alternatively, D sl architecture can consist of a copper connection from a residence to an intermediate node, called a street cabinet, which is connected with fiber to the central office. By installing street cabinets, telephone providers can offer Dsl service to areas farther than three miles from a central office.10 AT&T and Verizon are the main providers of D sl service in the united s tates, with AT&T dominant in California. Cable infrastructure consists of a head-end, which forwards local Internet traffic to the wider Internet and serves thousands of homes, and opti- cal nodes, which are connected to the head-end with fiber-optic cable and to residences with coax- ial cable. The coaxial cable is the same infrastruc- ture that delivers cable television service. To pro- vide broadband, cable television providers upgrade their networks by adding nodes and moving them closer to residences, which in effect replaces some of the coaxial cable in the network with fiber. Cable companies also have to install equipment (analogous to the telephone D slAMs) that route and switch digital data, and they install amplifi- ers that improve the upstream data transmission.11 u nlike D sl, cable infrastructure is shared: The coaxial cable connects residences to the optical node in a loop, so that the bandwidth any resi- dence receives depends in part on the number of residences sharing the node. Thus, the local fixed costs of upgrading an area’s infrastructure to make cable broadband available can involve (1) moving optical nodes closer to residences, (2) building new optical nodes, (3) upgrading the upstream path, and (4) installing equipment at the head-end.12 Comcast and Time Warner are the main providers of cable broadband service in California. These fi xed costs mean that cable and D sl are more profitable in areas where the costs can be spread over more subscrib- ers. A reas that are higher den- sity or higher income or both tend to be more profitable. In a higher-income neighbor- hood, more residents are likely to adopt broadband, so provid- ers seeking to make broadband available in the most profitable areas first would choose higher- income areas; this effect is mag- nified for cable providers, since upgrading their networks offers new potential revenue from both Because of high fixed costs and technological limitations, some cities or neighborhoods are more profitable for broadband providers to serve than others. broadband and digital television service. High- density neighborhoods are not only more likely to be within three miles of a telephone central of fice for Dsl service, they also can be served at lower installation costs per subscriber.13 For all of these reasons, broadband availabil- ity should differ geographically according to aver- age income and density. There might also be geo- graphic differences in availability that are unique to a particular provider: Most areas in the united s tates are served by a dominant telephone pro - vider and a dominant cable provider, and each can make different strategic decisions about when to introduce broadband service to their regions. The age and physical condition of existing telephone or cable infrastructure can also affect the cost of introducing broadband. In addition to these two wireline technologies, two wireless broadband technologies are begin- ning to be used.14 o ne is satellite, which although available nearly everywhere in the united states, offers a slower speed and lower reliability for a higher monthly price than either cable or D sl. 15 In prac- tice, satellite broadband appeals to consumers only where D sl and cable are unavailable and so does not actually compete with them.16 The other wireless broadband technology is Wi-Fi (the technical term is 802.11x), which offers high speed within a very short distance of a base station. Wi-Fi is commonly used in conjunc- tion with D sl or cable to make broadband access available wirelessly within a home, office, café, or public space. The cost of setting up base stations and antennas to provide Wi-Fi coverage is much lower than upgrading or building wireline infra- structures such as D sl, cable, or fiber-to-the-home (FTTH). In a handful of locations, Wi-Fi offers citywide public access to the Internet and actually competes with cable and D sl. This new wave of municipal Wi-Fi initiatives is discussed below as an example of broadband policy. California Economic Policy Broadband for All? 6 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Current Broadband Policies— Federal, State, Local F ederal, state, and local governments all play a role in shaping the availability and adop- tion of broadband. The federal government, through Congress, the executive, and the supreme Court, makes the most important regulatory deci- sions, whereas state and local governments play a larger role in subsidizing and in some cases directly providing broadband services. At the federal level, the FCC regulates tele- communications. The 1996 Telecommunications Act, the 2005 supreme Court decision in National Cable & Telecommunications Association et al. v. Brand X Internet Services et al., and related FCC rulings have created the regulatory framework that exists today. under this framework, telephone, wireless, television, and Internet providers are able to “compete in any market against any other” (FCC web site), but broadband providers are not required to give competitors wholesale access to their infrastructures so that competitors can resell services to consumers.17 The FCC also shapes broadband policy by placing conditions on merg- ers between broadband providers and by allocating wireless spectrum.18 Finally, the FCC administers the universal service requirement, which guaran- tees that even the most remote areas have telephone service.19 s tates and local governments are still left with important elements of broadband regulation, one being control of the physical development of infra- structure. Broadband networks typically follow public rights-of-way such as roads or rail tracks, and deploying broadband infrastructure costs providers less when it occurs in tandem with pub- lic works projects or when done simultaneously by multiple broadband providers. In California, improving rights-of-way access for broadband deployment is the most prominent element of Governor Arnold schwarzenegger’s 2006 Execu - tive order on “Expanding Broadband Access and u sage in California,” which created a California Federal, state, and local governments all play a role in shaping the availability and adoption of broadband. California Economic Policy Broadband for All? 6 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 7 Text Box 2. Will New Internet Access Technologies Overcome the Digital Divide? The economics of DSL and cable result in greater geographic differences in availability than those of dial-up access do. Will the next generation of technologies have the same effect or will they widen or narrow the digital divide? The most promising next-wave broadband technol - ogies are FT TH and WiMax.20 FT TH extends fiber-optic cable—which already connects the Internet all the way to the telephone company central of fices and the cable pro- viders’ optical nodes—the rest of the way to homes. Fiber would, in ef fect, replace the copper wiring and coaxial cable in use today. Because the capacity of fiber is far greater than that of either copper or coaxial cable, users would access much higher speeds downstream and upstream—potentially into the gigabit-per-second (gbps) range, hundreds of times faster than today’s fastest cable or DSL ser vices.21 Recent policy reforms at the state and federal levels have reduced some of the regulator y challenges to deliv- ering cable television and these give telephone companies a stronger incentive to build FT TH networks.22 Telephone companies are interested in FT TH as much for its abil- ity to deliver television ser vice as to deliver high-speed Internet ser vice; this is so that they can compete fully with cable providers, who can of fer television, Internet, and telephony (using voice-over-IP [Internet protocol]) over their existing networks.23 However, current adoption in California is minimal. SureWest, a regional telecom ser vice provider, has over 20,000 FT TH subscribers in the Sacramento area.24 AT&T, the dominant local phone provider, has not announced plans for any major FT TH deployment. Fiber costs more than DSL or cable, and it requires replacement of the existing connections to customers’ homes, so its roll-out proceeds neighborhood by neighbor- hood. Multiunit dwellings, dense areas, and new develop- ments cost less to wire with fiber than other areas, so there is a strong possibility of a future, persistent digital divide in FTTH availability; at least as important is that only some telecom companies are considering FTTH. In fact, a study prepared for San Francisco’s evaluation of the feasibility of a municipally built, owned, and operated fiber-optic network argues that the city is already on the losing side of the FTTH digital divide because Verizon, the company deploying most large FTTH projects in the country, is not the dominant telephone provider in San Francisco.25 The other promising next-wave technology is WiMax, a wireless technology. WiMax of fers Internet connectiv- ity over a range of one to 30 miles from a transmission tower; this compares to the hundreds of feet that are Wi- Fi’s limit. The downstream and upstream bandwidth of WiMax depends on the number of simultaneous users, but speeds could rival those of DSL and cable. WiMax is not yet being used for broad-based Internet access. Theo- retically, the fixed costs of WiMax deployment should be much lower per subscriber than costs for wireline tech- nologies, because the infrastructure consists of widely spaced antennas, not extensive wiring. The wide range of WiMax signals could also bring high-speed ser vice to harder-to-reach rural areas. Next-wave technologies are no guarantee of over - coming today’s digital divide. In fact, the high fixed costs of FT TH make it likely that some areas will receive ser- vice long af ter others do. Fur thermore, with the devel- opment of faster access technologies, expectations about adequate ser vice ratchet upward. This is not only because the digital divide refers to relative dif ferences, not absolute levels. It is also because online applications are designed for users’ current bandwidth; as typical resi- dential bandwidth increases, online applications incorpo- rate more bandwidth-hungr y content (such as video and interactivity), and access technologies that were once adequate cease to be so. California Economic Policy Broadband for All? 8 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A silicon Valley. Major initiatives are under negotia- tion for san Francisco, for the greater sacramento region, and for a silicon Valley–wide network; los Angeles has also announced a citywide initiative.31 Most of these call for low-cost or free wireless access, provided by partners such as Google, Earth- l ink, and MetroFi, supported by subscriptions or advertising, with little financial investment by the public sector. of the 58 initiatives, all but four are in densely populated areas of the Bay Area, south - ern California, or greater sacramento. Text box 3 provides more detail on the justifications and chal- lenges of municipal wireless. California’s broadband policy can be summed up as follows: First, subsidies focus more on insti- tutional access to broadband than residential access to broadband. second, municipal Wi-Fi initiatives are widespread and are concentrated in the densely populated parts of the state. Third, policy is geared toward broadband access, not computer ownership. Measuring Broadband T rying to measure the extent of the digital divide is challenging. Publicly available household surveys on broadband adop- tion, Internet access, and computer ownership are inadequate for studying recent trends in Califor- nia. For example, the federal Current Population Survey last included technology questions in 2003, and there are no plans to do so again. The Pew Internet & American life Project surveys house - holds about technology adoption annually or more frequently, but with only 4,000 respondents nationally, the sample is too small to draw con- clusions about California.32 A proprietary survey, the Technographics Benchmark conducted by For- rester, a technology research and consulting firm, is used for this analysis. Forrester annually surveys 60,000 –100,000 households about their technol- ogy adoption and behaviors.33 Measuring the divide in broadband availabil- ity is more challenging than measuring broadband adoption. Broadband providers treat service avail- Broadband Task Force composed of public and private stakeholders to coordinate efforts to raise broadband adoption and identify ways to fund new technology investments.26 s tates and localities also play a large role through subsidizing and directly providing broad- band. California’s primary broadband subsidy program is the Teleconnect Fund, which pays half the cost of Internet access for qualified schools, libraries, community organizations, and other nonprofits. Funded from a statewide fee on tele- phone service, the fund’s 2006 –2007 fiscal year budget is $22 million.27 In addition, the Califor- nia Public utilities Commission (CP uC) recently created the California Emerging Technology Fund, an independent nonprofit foundation to be funded with $60 million over five years from AT&T and Verizon as conditions of their respective mergers with sBC and MCI. The fund’s mission is “achiev - ing ubiquitous access to broadband and advanced services in California, particularly in underserved communities through the use of existing and emerg- ing technologies,” although specific strategies have not yet been selected.28 Whereas state governments have focused on subsidies to encourage adoption, more and more localities are attempting to provide broadband directly, both by themselves and in partnership with private companies. In the late 1990s, a few localities across the country built fiber-optic net- works.29 These early projects often involved public owner- ship of networks and were in direct response to the perceived lack of service provision by the phone and cable companies.30 In the past couple of years, many localities have turned to Wi-Fi as a wireless standard and a way to bring broadband service at low or no cost to a wide area. In Cal- ifornia, 58 localities have Wi-Fi initiatives under way. service is operational in Anaheim, the san Diego County tribal nations, and several cities in Whereas state govern- ments have focused on subsidies to encourage adoption, more and more localities are attempting to provide broadband directly, both by them- selves and in partnership with private companies. California Economic Policy Broadband for All? 8 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 9 Text Box 3. Local Wi-Fi Initiatives As of early 2007, numerous localities are developing wireless broadband networks using Wi-Fi technology to ser ve entire regions. In contrast to earlier ef for ts, most municipal Wi-Fi plans and deployments today call for at least par tial ownership and operation by the private sec- tor. So phone and cable companies, af ter fighting earlier attempts at direct public provision, are instead par tner- ing with local governments.34 The best-known are in Phil- adelphia, where a wireless network run by Ear thLink is operational, and in San Francisco, which is still negotiat- ing with Ear thLink and Google; across the countr y there are hundreds of others under way.35 Many arguments over municipal wireless rest on technical issues specific to Wi-Fi, which was not designed to provide city wide coverage but to bring it to a build- ing, park, or other small area. Wi-Fi transmits signals over relatively shor t distances (up to 30 meters indoors, 450 meters outdoors) and is the technology behind pub- lic “hotspots” and home networks. Municipal networks extend the capabilities of Wi-Fi by using multiple trans- ceiver sites that collectively cover a large area. The first technical concern about Wi-Fi is that it is unclear how far such a Wi-Fi signal can reach indoors, and users might need to install range-extending equipment. Second, new wireless standards such as WiMax that can transmit sig- nals much far ther could make Wi-Fi obsolete. Third, a city wide Wi-Fi signal could interfere with existing Wi-Fi hotspots.36 There is also controversy on social and econom - ic grounds. In San Francisco, the par tnership calls for Ear thLink to provide paid access and for Google to pro- vide free, adver tising-suppor ted access; Google’s model, which uses tracking cookies that customize adver tising, has raised privacy concerns.37 In some cities, the busi- ness model itself is being debated: Should wireless be adver tising-suppor ted (and free), or subscription-based, or a hybrid?38 Some are even questioning the public-pri- vate par tnership model: as of May 2007, San Francisco’s Board of Super visors has delayed finalizing the agree- ment with Ear thLink and Google to assess a city-owned and -operated alternative. Despite these controversies, there is broad consensus that the benefits of municipal Wi-Fi include (1) narrowing the digital divide and (2) facilitating online activities that are socially desirable or economically productive. Large cities such as Philadelphia and San Francisco focus most on the digital divide among residents and on bringing free or low-cost access to ever yone. Philadelphia, for in- stance, negotiated with Ear thLink to provide broadband at a lower cost to lower-income residents.39 Large cities also want to encourage socially desirable online behav- iors related to health care, education, and employment.4 0 Promoters of Silicon Valley’s request for proposal for a wireless network, however, emphasize the digital divide among businesses. Some are out of reach of both DSL and cable providers, they argue, and desirable online activities include those that improve business develop- ment, government ser vices, and public safety.41 Still unknown is how much demand there is among residents, businesses, and visitors for municipal Wi-Fi. Municipal Wi-Fi competes with existing fixed-wire con- nections in homes and workplaces; it also competes with data ser vices of fered by mobile phone providers, which of fer Internet connectivity to mobile phones and to spe- cially equipped laptops. Taipei has one of the world’s most extensive wireless networks, reaching 90 percent of the city’s 2.6 million people, but six months af ter its launch in early 2006, only 40,000 residents had subscribed.42 California Economic Policy Broadband for All? 10 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A ability maps as sensitive competitive information, and regulators have not required that provid- ers make these maps public. Although individual consumers can check availability at an individual address on most providers’ web sites, anecdotal evidence suggests that this information is often inaccurate or incomplete.43 The only widely used data on broadband avail- ability come from FCC surveys of broadband pro- viders. The FCC publishes a count of the number of providers with at least one subscriber in each zip code nationally.44 This method overstates the level of availability because broadband may be available in only some parts of a zip code. using this measure could also under- state geographic differences in broadband availability if provid- ers make broadband available only in richer or denser portions of a zip code. recent FCC data suggest implausibly high levels of broadband availability, ren- dering the data misleading for assessing the digital divide. The FCC reports that in December 2005, 99.8 percent of the u.s . population lived in a zip code where broadband was available (see Table 1). In the most sparsely populated tenth of zip codes, 96.2 percent of the population lived in zip codes where broadband was available, and in the poorest tenth of zip codes, 99.4 percent of the population lived in zip codes where broadband was available. nevertheless, these FCC data are the basis for most policy studies and academic work about broadband availability. In its most recent report on broadband deployment in the united states in 2004, the FCC concludes that broadband is nearly univer- sally available in urban areas and “significant prog- ress is being made towards ubiquitous availability of advanced services in rural areas.”45 The California Public utilities Commission (2006) also relies on the FCC’s zip code data for its maps of broadband availability, which show at least one broadband pro- vider offering service in every zip code in Califor- nia.46 s everal academic studies rely on the FCC data as well, although one notes that “the FCC count of high speed line providers within a zip code may seri- ously misrepresent competitive options available to the totality of residents within that zip code … [but] there is no practical alternative to using the FCC data in assessing broadband availability.”47 r ather than rely on FCC data, this analy - sis infers broadband availability by examining Table 1. U.S. Broadband Availability, FCC Measure Population in Zip Codes with At Least One High-Speed Subscribera (%) Population of Lowest-Density Zip Codes (bot tom decile) in Zip Codes with At Least One High-Speed Subscriber (%) Population of Lowest-Income Zip Codes (bot tom-decile) in Zip Codes with At Least One High-Speed Subscriber (%) December 2 0 0 09 6.44 9.991.5 December 2 0 01 97.8b67.995.1 December 2 0 0 29 9.18 0.997.5 December 2 0 0 39 9.58 8.99 8.6 December 2 0 0 49 9.691.89 9.0 December 2 0 059 9.89 6.29 9.4 source: Federal Communications Commission (2006), Tables 18 and 19. aBased on FCC data and the author’s calculations.bCompared to an online Forrester survey in autumn 2001, in which 64 percent of online respondents reported broadband availability where they live. Rather than rely on FCC data, this analysis infers broadband availability by examining the relation- ship between location and broadband adoption at the individual level, controlling for numerous individual characteristics. growth in broadband adoption is not yet leveling off. Throughout this period, the level of broadband adoption in California has been above the national rate. Most recently, 47 percent of California house- holds had broadband, eight points higher than the overall national level. There are several possible explanations for this. o ne is that California residents have more favor - able demographics for broadband adoption—that is, they are richer, more educated, and so on—than Americans generally, and these demographic factors influence broadband adoption. A second possibility is that broadband is more widely available in California because the density of population and income levels make it more prof- itable to offer broadband here than elsewhere in the country. A third possibility is that another California-specific factor—such as state policies or decisions by AT&T, the dominant local tele- phone provider in California— contributed to faster deployment and therefore higher adoption. These three possibilities are not mutually exclusive. California Economic Policy Broadband for All? 10 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 11 Looking just at the 68 percent of U.S. house- holds with Internet access (broadband or dial-up) at home, more than half have broadband. the relationship between location and broadband adoption at the individual level, controlling for numerous individual characteristics.48 This tech- nique, in essence, compares people who are identi- cal in observable demographics, such as income, age, education, size of household, age of children, and so on, but who live in different types of zip codes—for example, a middle-income urban zip code and a middle-income rural zip code.49 Because their individual demographics are the same, they are assumed to have similar underlying demand for broadband, and any difference in whether they actually have broadband is interpreted as a differ- ence in the availability of broadband in the differ- ent types of zip codes.50 The key location charac- teristics are zip code log median household income and zip code log population density.51 Measuring the digital divide in complementary skills presents further difficulties. self-perceptions of skills are harder to elicit and interpret, so research on complementary skills often relies on direct obser- vation and usability testing, which is expensive to conduct on a large scale. Furthermore, as technol- ogy changes, particular skills rise or fall in impor- tance, making measurement and interpretation even more difficult.52 n or do any of the main data sources on access or availability, including Forrester, solicit detailed information on complementary skills, so this aspect of the digital divide is beyond the scope of this report. Broadband Availability and Adoption in California O ur first research question is whether Cali- fornia leads or lags the nation in broad- band adoption. nationally, residential broadband adoption has grown rapidly, from 7 percent of households at the end of 2000 to 39 per- cent in 2005 (see Figure 1).53 l ooking just at the 68 percent of u.s . households with Internet access (broadband or dial-up) at home, more than half have broadband. During 2005, broadband adop- tion grew from 28 percent to 39 percent, so the Figure 1. Broadband Adoption in the United States and California Source: Forrester Research and author’s calculationsNote: Data based on survey conducted by mail in English only. Percentage 200420032002200120002005 45 40 35 30 25 20 15 10 5 50 0 California United States 47% 39% 10%7% 15% 18% 28% 10% 17% 23% 24% California Economic Policy Broadband for All? 12 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A The results of this analysis suggest that about half of the gap between broadband adoption in California and that in the rest of the nation is due to California’s favorable individual demographics. To assess the importance of each, regression anal- ysis can identify the separate relationships of individ- ual demographics and location characteristics such as neighborhood income and density on broadband adoption. The results of this analysis suggest that about half of the gap between broadband adoption in California and that in the rest of the nation is due to California’s favorable individual demograph- ics.54 l ocation characteristics that affect providers’ decisions to make broadband available—neighbor- hood income and population den- sity—account for the other half of the gap.55 Because individual demographics and location char- acteristics account for the entire difference between broadband adoption in California and that in the rest of the country, we infer that other factors unique to California did not collectively change the level of broadband adoption.56 The second research question is whether there are inequalities in broadband availability within California. As noted above, this report infers avail- ability from the relationship between location characteristics and broadband adoption. Another regression analysis reveals that broadband adop- tion in California is significantly higher in higher- income and denser zip codes, even after controlling for individual demographics.57 That is, people of the same age, race, income, and so on are more likely to adopt broadband if they live in a richer or denser area.58 The effect of location on avail- ability means that broadband adoption differs con- siderably across regions within California, some of which are much richer and denser than others. Broadband adoption ranges from under 30 percent in the north of the state and in the sierras to just over 50 percent in the Bay Area and greater los Angeles (see Table 2).59 The differences in Inter- net access (broadband and dial-up combined) and computer ownership across regions are smaller: Internet access ranges from 61 percent to 76 per- cent and computer ownership ranges only between 73 percent and 83 percent. More than half the dif- ferences in broadband adoption between regions remain even after controlling for individual demo- graphics, suggesting that inequalities in broadband availability across regions account for much of the differences in broadband adoption. In contrast, the differences between regions in overall Internet access and computer ownership are due to differ- ences in individual demographic characteristics across those regions.60 The third research goal is to understand the inequalities in broadband adoption in California apart from availability. Although it is obvious (as we confirm below) that broadband adoption rates are higher among richer households, two impor- tant questions about income and broadband adop- tion whose answers are not obvious are (1) do race and ethnicity affect broadband adoption, holding income constant, and (2) does income affect broad- band adoption because income also affects com- puter ownership, which is essentially a prerequisite for broadband adoption?61 Broadband adoption, Internet adoption, and computer ownership all vary by income in Cali- fornia. Among households with incomes over $100,000, 68 percent have broadband, compared to 49 percent of households with incomes between $50,000 and $75,000 and compared to 24 percent of households with incomes under $25,000 (see Table 3). The gaps for Internet adoption (broad- band and dial-up combined) and computer owner- ship across income groups are also considerable. looking across racial and ethnic groups, it is important to keep in mind that Forrester conducted this survey in English only. These data therefore represent people with very high English proficiency and exclude a significant share of California’s Hispanic population.62 Disparities in broadband adoption across racial and ethnic groups are less wide than across income groups but they are still apparent, with 46 percent of both non-Hispanic whites and English-proficient Hispanics having broadband, compared with 63 percent of Asian Americans and only 36 percent of African Ameri- cans.63 s ince income among non-Hispanic whites is English and spanish, broadband adoption is 50 percent among Hispanics who speak English as much or more than spanish at home and responded to the sur- vey in English. Broadband adop- tion is only 20 percent among Hispanics who speak spanish more than English at home or responded to the survey in span - ish. Broadband adoption among all California Hispanics is 29 percent, so looking only at His- California Economic Policy Broadband for All? 12 P u Bl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 13 higher than income among Hispanics, and income and broadband adoption are strongly related, why do both groups have the same level of broadband adoption in California? relative to non-Hispanic whites, Hispanics on average have larger house- holds and are younger, both of which are positively correlated with broadband adoption. other research shows that Hispanics in Cali - fornia who prefer speaking spanish at home have much lower broadband adoption than English- speaking Hispanics. According to the March 2007 PPIC statewide survey, conducted by telephone in Table 2. Broadband, Internet, and Computer Ownership, by California Region, 2005 Broadband (%) Online (broadband or dial-up) (%)Computer (%)Number Nor thern California2 96 37 7161 Nor thern Sacramento Valley2 8697916 3 Greater Sacramento4 4768 3518 San Francisco Bay Area51748 01, 3 3 5 Nor thern San Joaquin Valley3 66774241 Southern San Joaquin Valley3 56 4733 8 7 Central Sierra2161765 4 Central Coast4 8768 0176 Greater Los Angeles5276811,9 4 9 Inland Empire4 573826 6 0 San Diego Border4 8727872 2 s ources: Forrester r esearch and the author’s calculations. n otes: Counties in each region are listed in the web-only appendix at www.ppic.org /content /other/707J KEP_web_ appendix.pdf. Data are based on a survey conducted by mail in English only. Table 3. Broadband, Internet, and Computer Ownership in California, by Income, 2005 Household Income ($ 1,000s)Broadband (%) Online (broadband or dial-up) (%)Computer (%)Number < 2 5244 85 81,167 2 5 – 4 94 0707 71,573 5 0 – 694 9788 61,14 6 70 – 9 9598 7911, 32 0 10 0 +6 88 99 31, 3 82 sources: Forrester r esearch and the author’s calculations. n ote: Data are based on a survey conducted by mail in English only. Broadband adoption ranges from under 30 percent in the north of the state and in the Sierras to just over 50 percent in the Bay Area and greater Los Angeles. California Economic Policy Broadband for All? 14 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A panics with high English proficiency overstates the level of broadband adoption among Hispanics.64 Holding other individual demographics con- stant, there is a statistically significant difference for broadband adoption among Hispanics and African Americans relative to whites. English- preferring Hispanics are 6 percent less likely to have broadband at home than non-Hispanic whites; the difference for African Americans is 10 percent.65 r elative to California’s overall broad- band adoption of 47 percent, these gaps are large. Moreover, for Hispanics this is the gap between adoption among non-Hispanic whites and Hispan- ics with high English proficiency. The racial gaps in broadband adoption, however, cannot be explained by provid- ers’ failing to offer service in minority neighborhoods because there was no statistically signifi- cant effect of the racial composi- tion of a zip code on broadband adoption. The other important fac- tor about the digital divide in broadband adoption is the role of computer ownership. A home computer is for the most part a prerequisite for broadband adoption.66 If the relationship between income and broadband adoption reflects lower computer own- ership by poorer households, then making broad- band more widely available at low cost might have little effect on broadband adoption. looking across income groups, computer ownership ranges from 58 percent among households with income under $25,000 to 93 percent among households with income $100,000 and above. regression analysis helps assess how much this gap in computer own- ership contributes to the digital divide in broad- band adoption. Controlling for other demograph- ics, raising log income by one standard deviation increases the likelihood of computer ownership by 7 percent.67 Then, looking only at computer own- ers and again controlling for other demographics, raising income by the same amount increases the likelihood of broadband adoption by 10 percent.68 Both relationships are statistically significant: The effect of income on computer ownership means that there is a digital divide in computer owner- ship, but the effect of income on broadband among computer owners means that, even among com- puter owners, there is a digital divide in broadband adoption. Conclusions and Recommendations W ithin California, there are digital divides in both broadband availability and broadband adoption. Broadband is more widely available in higher-density residen- tial areas and urban regions than in lower-density areas and rural regions; the level of neighborhood income also affects availability. These inequali- ties arise, in part, from the technical features and economic realities of broadband provision. The digital divide in adoption encompasses not only broadband but computer ownership as well, and income and race and ethnicity both affect broad- band adoption in California. our findings have important implications for broadband policy. If closing gaps in broadband availability is a policy goal, raising availability in rural areas should be the top priority. The Califor- nia Emerging Technology Fund should focus first on deployment in rural areas, and the Broadband Task Force could help identify clearly the barriers to providers’ offering service in rural areas. Eas- ing the access to rights-of-way might facilitate rural broadband deployment, but the economics of broadband make rural areas costly to serve even in the absence of any regulatory or permitting fac- tors, so the state could consider direct subsidies to providers serving rural areas. Although broadband availability lags in lower-income areas as well, den- sity has a greater effect on availability; furthermore, there is no evidence that broadband availability is lower in neighborhoods with higher concentrations of Hispanics and African Americans. Holding other individual demographics constant . . . English-preferring Hispanics are 6 percent less likely to have broadband at home than non-Hispanic whites; the difference for African Americans is 10 percent. California Economic Policy Broadband for All? 14 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 15 To close gaps in broadband adoption, the focus should be broadened to include racial and ethnic and income groups with lower adoption rates, including those living in neighborhoods where broadband service is already available. Improving technology literacy may also be essential. Munici- pal Wi-Fi initiatives will help meet these goals so long as they provide broadband at lower cost than Dsl and cable providers do. recent initiatives sug - gest that they will: Google’s proposal to provide free, ad-supported service in san Francisco, and the prevailing rate of about $20 per month that Earth link charges for its Wi-Fi service, are two examples. Municipalities need not resort to pub- lic ownership or operation of Wi-Fi networks to achieve these public benefits. Public involvement should be reserved for the investments that the pri- vate sector is less likely to make, such as helping raise technology literacy and improving computer access among disadvantaged residents, both of which are among the “digital inclusion” goals of san Francisco’s Wi-Fi initiative. 69 Ironically, because the Wi-Fi networks that are operational or under consideration are over- whelmingly in the urban parts of the state, they are likely to widen rather than close the gap in avail- ability between urban and rural areas. Policymak- ers, therefore, should think less in terms of closing the gap between urban and rural availability, lest they hold back urban initiatives, and more in terms of raising the absolute level of rural availability. n ew technologies, such as FTTH, are also likely to benefit urban areas more than rural areas, so the urban-rural gap is likely to persist even if policy efforts successfully raise the level of rural broad- band availability. Two broader questions remain to be researched and answered. First, is it more desirable from a cost-benefit perspective to raise the degree of broadband adoption for those with less access than it is to overcome other inequalities that Californians face? on one hand, the economic and social benefits of broadband are unknown and are largely taken as an article of faith; on the other, the costs to the public sector of raising broadband adoption can be low if the private sector bears the cost, as is the case with most municipal Wi-Fi initiatives. sec - ond, will the development of future Internet access technolo- gies such as WiMax or FTTH mitigate or aggravate today’s inequalities in tech- nology access? Both of these questions should be considered— even if they cannot be answered— as California’s state and local governments weigh various strategies for raising broadband adoption and closing the digital divide. v If closing gaps in broadband availability is a policy goal, raising availability in rural areas should be the top priority. To close gaps in broad- band adoption, the focus should be broadened to include racial and ethnic and income groups with lower adoption rates. California Economic Policy Broadband for All? 16 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Notes 1 The California Public utilities Commission (2005) cites California’s #14 ranking among states (according to the 2003 Tech net survey) as an important indicator of wheth-er the state is “maintaining its lead in broadband usage.” In the California Public utilities Code, the main broad-band policy objectives are assuring “continued affordabili-ty and widespread availability” of broadband and promot-ing “economic growth, job creation, and substantial social benefits.” It is hard to see how the “substantial social ben-efits” depend on California’s broadband adoption relative to that of other states, rather than the absolute level, but it is plausible that the relative ranking could affect economic growth and job creation if businesses consider broadband availability and adoption a factor in deciding where to open, expand, or move their operations. 2 o nly limited research has been completed on the effects of broadband. An overview of the academic literature on the effects of broadband and Internet usage is presented in Text Box 1. 3 California Public utilities Commission (2005) summarizes several studies on broadband and economic development. 4 Gunkel (2003). 5 see Morton, Zettelmeyer, and s ilva-risso (2003). 6 Dial-up Internet access, in contrast to cable and D sl, is nearly ubiquitous in the united states and never had a long period of geographic disparities in availability. Mak-ing dial-up available does not involve a high fixed cost: Dial-up Internet service providers (I sPs) need to establish a point-of-presence (P oP) in an area by making a local phone number available as a dial-in number, and although this involves adding some switching equipment in the tele-phone company central office, the infrastructure invest-ment is minimal. Downes and Greenstein (2002) docu-ment that dial-up Internet service spread quickly to even the most rural counties: In 1997, 99 percent of the u.s . population lived in counties with at least one I sP, and 92 percent lived in counties with seven or more I sPs. There are scale economies in having a P oP, but the costs to upgrade infrastructure for broadband service are much larger (Greenstein and Prince, 2006). 7 D sl refers to a family of technologies, technically de- scribed as xD sl. residential D sl is most often AD sl (asymmetric D sl), and higher-capacity variations include s D sl and V D sl. 8 Many smaller businesses subscribe to D sl as well. larger businesses rely on higher-bandwidth connections such as T1 and T3 lines. Cable broadband is almost exclusively a residential service. 9 The description of Internet infrastructure and the process for upgrading networks relies on Federal Communications Commission (2000) and Corning (2001, 2005). 10 The website www.dslreports.com/prequal/distance offers estimates of distance between a residential address and the nearest central office and maps of central office locations. 11 Traditional cable television is a one-way communica-tion: Video is broadcast from the cable head-end to the residence. Internet service, as well as digital cable service, is two-way communication, which requires a different “upstream” infrastructure. 12 Cable providers might have to make additional infra-structure investments as broadband adoption rises, con-gestion increases, and bandwidth speeds fall. 13 High residential density is one reason why Korea, Hong Kong, and Japan have higher broadband adoption and higher speeds of service available than the united states does. 14 In addition to satellite and Wi-Fi, there is also WiMax, which is discussed in Text Box 2. 15 satellite also has a higher upfront equipment cost. Hughes net basic service costs $59.99 per month for speeds up to 700 kbps plus a $400 equipment and installation fee. see the website go.gethughesnet.com. 16 s ee “With a Dish, Broadband Goes r ural” (2006). 17 The laws and rulings that make up the federal regula-tion of broadband have involved the sometimes competing goals of (1) maximizing competition among broadband providers, (2) not reducing the incentives that broadband providers have to invest in infrastructure and new applica- tions, and (3) treating D sl and cable broadband services evenly despite the fact that telephone companies as “com-mon carriers” have traditionally been subject to much stricter federal regulation than cable companies have. 18 In 2005, the FCC required that AT&T offer “naked D sl ,” under which consumers can subscribe to AT&T’s D sl service without subscribing to AT&T’s local telephone service, as a condition of its merger with sBC. Then, in late 2006, the FCC required specific pricing for naked D sl as a condition of its merger with Bell south. see “AT&T to o ffer ‘ naked D sl’ for Far l ess Than Before” (2007). 19 A short explanation of universal service is available at the FCC’s website, www.fcc.gov/cgb/consumerfacts/ universalservice.html. universal service applies only to telephone, not Internet, service. However, because D sl uses the telephone network, the cost of providing D sl in remote areas is lower than if the telephone network did not cover those areas. In contrast, cable companies face no universal service requirement. As a result, Dsl adoption is higher than cable broadband adoption in very low-density rural areas. 20 Fiber-to-the-home is one technology in a class of FTTx technologies: They differ in whether the fiber network extends all the way to the individual home, or to the curb (FTTC), or to a multiunit building (FTTB). Fiber networks include those that have dedicated fiber running between the California Economic Policy Broadband for All? 16 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 17 customer and the service provider (point-to-point networks) or have shared fiber (as in a “passive optical network”). 21 Another emerging wireline technology is Broadband over Powerline (BP l), which uses the electric power network and home electrical wiring and outlets to carry Internet data. unlike FTTH, BP l speeds are similar to those of D sl and cable (see the website www.fcc.gov/cgb/consumerfacts/highspeedinternet.html). 22 In 2006, California passed a law creating statewide video franchises, which allow television providers to apply for a single statewide franchise rather than multiple local franchises as cable companies have traditionally had to do. Also in 2006, the FCC put time limits on the local franchise negotiation process and on franchise fees; it also curtailed unreasonable “build-out” requirements, which specify how quickly television providers are required to offer service to the entire franchise area. see the C nET.com article at http://news.com.com / FCC+adopts+relief+for+ telecom+companies+planning+T V+offerings/2100 -1036 _3- 6145184.html. 23 Verizon’s television-over-IP service is called Fi os. AT&T delivers its television service, u-Verse, by extending its fi-ber network closer to, but not all the way to, residences and then delivering content over the last portion of the network using a higher-capacity version of D sl. see the C n ET.com article at http://news.com.com /AT T+to+ramp+ up+IPT Vs+expansion /2100 -1037_ 3- 6153354.html. 24 s ee sureWest’s 2006 Q3 financial statement, at www.su r w.com /med ia _ relat ions /press /releases /ea r n i ngs /pd f /Q3_06financials_p6.pdf. 25 n one of san Francisco’s broadband providers plan to deploy FTTH to san Francisco beyond limited trials, and these providers’ future plans are “not comparable” to FTTH (Columbia Telecommunications Corporation, 2007, p. 2). Verizon, in contrast, plans to spend $20 billion over several years to make fiber-to-the-home high-speed services avail-able to 60 –70 percent of customers in its current service area, which includes some areas in California, according to “Verizon’s Fios services Build Momentum” (2006). 26 The Executive order designates the Business, Transpor-tation, and Housing Agency as the lead agency for most items, highlighting the importance of infrastructure devel- opment to the goal of broadband access. on many telecom policy issues, including rate regulation, the Public utilities Commission has responsibility. 27 California Public utilities Commission (2005, p. 62) has more details of the program and also of the similar federal e- rate program. 28 s ee the fund’s website at www.cetfund.org. 29 In Tacoma, Washington, the municipal power utility, City light, built, owns, and operates a fiber-optic network that delivers Internet and television. see “Cities Deliver Broadband for l ess” (2003). 30 Gillett, lehr, and osorio (2004) review many of these local broadband initiatives. 31 A full list of municipal wireless initiatives in the united s tates is available at www.muniwireless.org. status of ini-tiatives comes from the January 2007 summary. 32 The most recent Pew report on broadband adoption is based on a survey of 4,000 adults over the period February–April 2006. see www.pewinternet.org /pdfs/ PIP_ Broadband_trends2006.pdf. 33 Forrester’s annual Technographics Benchmark survey is conducted by mail, in English only; the samples are selected from national market research panels to be rep- resentative of u.s . households demographically and are weighted to correct for differences in response rates. For- rester has used the T ns (formerly nFo ) market research panel since 2001 and used nPD’s panel in earlier years. Forrester collects data in the 48 contiguous states and the District of Columbia but not in Alaska or Hawaii. some respondents participate in Forrester’s survey in multiple years: Kolko (2007) includes some longitudinal analysis from these data. Brown and Goolsbee (2002), Goolsbee (2000), Goolsbee and Klenow (2002), and Prince (2003) use Forrester’s data as well. The author was employed at Forrester from 2000 to 2005. 34 “Companies That Fought Cities on Wi-Fi, now rush to Join In” (2006). 35 A comprehensive list is available at www.muniwireless.com. 36 “Wi-Pie in the sky” (2006) reviews the technological and managerial challenges of citywide Wi-Fi networks. 37 “s ome Worries as san Francisco Goes Wireless” (2006). 38 Google provides free Wi-Fi in Mountain View, Cali-fornia, and MetroFi offers both free and paid Wi-Fi in Cupertino, sunnyvale, and santa Clara, California. see “ s.F. Picks Google Wi-Fi Team” (2006), and “Google Gives City Free Wi-Fi” (2006). 39 In Philadelphia, Earth link charges $21.95 per month generally and $9.95 per month for low-income households. 40 s ee, for instance, the literature from Wireless Philadelphia, available at www.phila.gov/wireless/briefing.html; from s an Francisco TechConnect, available at www.sfgov.org / site/techconnect_tf_index.asp; and from an interview with Houston’s chief information officer at www.govtech.net/digitalcommunities/story.php?id=98722. 41 “Public, Private Collaboration To Design silicon Val- ley Wide Wireless network” (2006). silicon Valley, unlike most large cities, has a lower share of low-income residents and more geographically dispersed businesses that aren’t well served by D sl or cable providers. California Economic Policy Broadband for All? 18 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 42 s ee “What If They Built an urban Wireless network and Hardly Anyone used It?” (2006). The article cites the competition from Wi-Fi connections, mobile phone data services, and fixed-wire broadband as a major reason for the lack of interest in paying for Taipei’s municipal Wi-Fi service. 43 A personal anecdote: over a two-day period in August 2006 of checking the AT&T website and calling several customer service departments to sign up for new D sl ser- vice, the author was told by different people that (1) D sl was unavailable at the address, (2) only a slower D sl ser -vice (up to 1.5 Mbps) was available at the address, (3) all speeds of Dsl service were available at the address, and (4) the address was not a valid address. 44 These counts include both providers who own the broad- band pipes (such as the dominant telephone and cable companies) and resellers. Although residential custom-ers account for most of the dominant telephone and cable companies’ subscribers, some resellers serve primarily business customers. 45 FCC (2004), p. 38. The FCC uses “advanced services” as a synonym for broadband. Two of the five FCC com-missioners dissented from the upbeat conclusions of the report, and both noted the inadequacy of using the one- subscriber-in-a-zip-code measure of availability. still, the main concern of both dissenters was the lower level of broadband adoption in the united states relative to that in Korea, Japan, and other countries, not the inequality of broadband availability within the united states. Further-more, not only do global broadband leaders have a higher percentage of households with broadband, average broad-band speeds are also much higher. 46 The CP uC emphasizes the inequality in the number of providers in zip codes, rather than using the data as evidence of ubiquitous access to broadband, and notes that “this rep-resentation does not depict the availability of broadband for every resident in each identified zip code area.” 47 Flamm (2006). Grubesic (2006), and Prieger (2003) also use FCC broadband data. 48 This section of the report draws heavily on Kolko (2007). Details on the methodology for inferring broadband avail-ability can be found there. 49 u nlike the CP s or Pew, Forrester includes respondents’ zip codes. 50 This approach raises two methodological concerns. First, location characteristics could be capturing unobserved indi- vidual characteristics. second, location itself could affect the demand for broadband if, for instance, people with less access to doctors or stores have greater demand for online health advice or online shopping. Kolko (2007) finds that location characteristics have a much smaller or insignifi-cant effect on technologies other than broadband and that location characteristics strongly influence the choice of cable versus D sl, which is much more plausibly due to availability differences than to location-driven differences in underlying demand for one broadband technology ver- sus the other. still, it is important to keep in mind that this approach measures availability as the relationship between location characteristics and adoption, which might also be capturing some effects in addition to availability. Further-more, this measure does not capture reasons other than zip code density and zip code income for variation in availabil-ity, such as zip-code-level topographical differences that might make some zip codes more expensive to serve. 51 This approach implicitly captures geographic differences in broadband prices. If there is more competition among broadband providers in higher-density areas, this could result in lower prices and therefore higher adoption. The notion of “availability” used here is not just a binary mea-sure but rather a continuous measure that captures dimen-sions of availability such as price and quality of service. 52 DiMaggio et al. (2004) review the academic literature on the digital divide in complementary skills and provide more detail on the challenges researchers face in studying it. 53 Forrester’s measure of broadband refers to “high-speed Internet access” and mentions D sl and cable modem as examples. In other questions in Forrester’s surveys that ask about specific technologies, broadband includes D sl, cable, fixed wireless, and satellite. 54 That is, the coefficient on the California dummy vari-able in a probit regression where broadband adoption is the dependent variable falls by about half when individual demographics are included. 55 That is, the coefficient on the California dummy variable in a probit regression where broadband adoption is the dependent variable is not significantly different from zero when individual demographics and zip code characteris-tics (log median income and log density) are included. 56 Because there could be many factors unique to Califor-nia that affect broadband adoption, one cannot conclude that they are irrelevant individually. For instance, if, hypo-thetically, AT&T had been unusually aggressive in rolling out broadband in California, but state policy made deploy-ment more difficult than in other states, these two effects could cancel each other out. Thus, we emphasize that fac-tors unique to California did not collectively have an effect on broadband adoption. 57 The web-only appendix www.ppic.org /content/other/707JKEP_web_appendix.pdf presents regression results for California and also for the united states. Higher-income and higher-density zip codes have higher broadband adop-tion, controlling for individual characteristics, both for California and the united states. The effect of income is smaller for California than for the united states, although statistically significant in both cases. The effect of density is similar for California and the u nited states. California Economic Policy Broadband for All? 18 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 19 58 These significant and positive relationships do not hold when Internet access (broadband or dial-up) is the depen-dent variable, or when computer ownership is the depen-dent variable. This is further evidence that the relation-ship between zip code characteristics and adoption is due to availability, since the availability of dial-up access and computer ownership do not vary geographically. relative to the overall broadband adoption level of 47 percent, liv-ing in a zip code with a log density one standard deviation (1.72) above the mean raises the likelihood of broadband adoption by six percentage points—an amount not only statistically significant but large in magnitude as well. 59 Table 2 shows the nine regions defined by the Economic s trategy Panel, but we have split the san Joaquin Valley into northern and southern regions, and we have split the south - ern California region into Greater los Angeles and Inland Empire. The distribution of Forrester’s respondents across these regions is similar to the population distribution. 60 That is, an F-test does not reject the hypothesis that the regional dummies all equal zero in a regression of Inter-net adoption (and, separately, computer ownership) on individual characteristics and regional dummies. This test does reject the hypothesis of insignificant regional dum-mies when broadband adoption is the dependent variable. 61 Although this section focuses on race, ethnicity, and in-come, it is also a goal of California policy to consider the digital divide between people with disabilities and people without. Forrester’s surveys do not ask about disability status. 62 The Forrester Technographics survey is long, detailed, and conducted by mail, so respondents need considerable English proficiency to complete it. Furthermore, both His-panics and African Americans are underrepresented in Forrester’s survey relative to their shares in the population reported by the Census. However, the income distribu-tions for these groups in Forrester’s survey are quite close to those in the March 2005 Current Population survey, so Forrester’s sample appears to be a reasonable enough representation for us to report results. 63 some non-Hispanic whites, African Americans, and Asian Americans also lack English proficiency, but the English proficiency requirement should affect the repre-sentativeness of the Hispanic sample most. 64 Because Hispanics preferring to speak spanish are a considerable share of California’s population, the overall level of broadband adoption in California rises from 55 percent among all Californians to 65 percent when these Hispanics are excluded, suggesting that the Forrester figure for English-speaking Californians would fall by up to 10 percentage points if non-English-speakers were also sur-veyed. These figures are higher than the Forrester figures for overall adoption in part because the PPIC survey was conducted 15 months after the Forrester survey. Also, the PPIC data, unlike Forrester data, show a gap in adoption between Hispanics preferring to speak English and whites, in part because a higher level of English proficiency would be required to complete Forrester’s mail survey in English than to participate in PPIC’s telephone survey in English. An earlier survey, the october 2003 Current Population Survey, reports that broadband adoption is 18 percent for California Hispanics who are not spanish-only versus only 2 percent for spanish-only Hispanics in California. Data from PPIC and the Current Population Survey are weighted using standard demographics to represent the population more accurately, but the figures reported here do not further adjust for demographics using a regression framework. 65 The regression underlying this finding also controls for zip code income and density. 66 Why might a household have broadband without a com- puter? some videogame consoles, digital video record-ers, and smartphones can connect directly to the Internet using cable or D sl. Households could also have broad-band access for use with an employer-provided computer. In California, 80 percent of households have a computer. Among computer-owning households, 58 percent have broadband at home. Among households with broadband, 98 percent own a computer. 67 r aising log income by one standard deviation at the mean of log income is equivalent to raising income from $48,000 to $114,000. The standard deviation of log income is .875. 68 These results are from (1) a probit regression of computer ownership on individual characteristics and (2) a probit regression of broadband adoption on individual and zip code characteristics, conditional on computer ownership. 69 Another strategy for increasing computer ownership is manufacturing and distributing extremely low-cost com- puters, which is the goal of the nonprofit one laptop Per Child foundation, www.laptop.org. Its focus is raising computer use among children in developing countries. California Economic Policy Broadband for All? 20 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A Federal Communications Commission, Availability of Advanced Telecommunications Capacity in the United States: Second Report to Congress, Washington, D.C., 2000. Federal Communications Commission, Availability of Advanced Telecommunications Capacity in the United States: Fourth Report to Congress, Washington, D.C., 2004. Federal Communications Commission, High Speed Ser-vices for Internet Access: Status as of December 31, 20 05, Washington, D.C., July 2006. Flamm, Kenneth, “Diagnosing the Disconnected: Where and Why Is Broadband Access unavailable in the u.s .?” unpublished draft, 2006. Gillett, sharon, William lehr, and Carlos osorio, “ local Government Broadband Initiatives,” Telecommunications Policy, Vol. 28, 2004, pp. 537–558. “Google Gives City Free Wi-Fi,” San Francisco Chronicle, August 16, 2006. Goolsbee, Austan, “In a World Without Borders: The Impact of Taxes on Internet Commerce,” Quarterly Jour-nal of Economics , Vol. 115, no. 2 , 2000, pp. 561–576. Goolsbee, Austan, and Peter Klenow, “Evidence on learn - ing and network Externalities in the Diffusion of Home Computers,” Journal of Law and Economics, Vol. 45, 2002 , pp. 317–343. Greenstein, shane, and Jeff Prince, “The Diffusion of the Internet and the Geography of the Digital Divide in the u nited states,” nBE r Working Paper #12182 , Cambridge, Massachusetts, 2006. Grubesic, Tony, “A spatial Taxonomy of Broadband r egions in the united states,” Information Economics and Policy, Vol. 18, 2006, pp. 423 – 448. Gunkel, David, “ second Thoughts: Toward a Critique of the Digital Divide,” New Media & Society, Vol. 5, 2003, pp. 499–522. Kolko, Jed, “Why should Governments support Broadband Adoption?” Public Policy Institute of California, Working Paper #2007.01, s an Francisco, California, 2007. Kuhn, Peter, and Mikal skuterud, “Internet Job search and unemployment Durations,” American Economic Review , Vol. 94, n o. 1, March 2004, pp. 218 –232. Morton, Fiona scott, Florian Zettelmeyer, and Jorge silva- r isso, “Internet Car retailing,” Journal of Industrial Economics , Vol. 49, n o. 4, 2001, pp. 501–519. Bibliography “AT&T to offer ‘ naked D sl’ for Far less Than Before,” San Francisco Chronicle, January 16, 2007. Brown, Jeffery, and Austan Goolsbee, “Does the Internet Make Markets More Competitive? Evidence from the life Insurance Industry,” Journal of Political Economy, Vol. 110, n o. 3, 2002 , pp. 481–507. California Public utilities Commission, “Broadband Deployment in California,” san Francisco, California, 2005. California Public utilities Commission, “Broadband Deployment in California,” san Francisco, California, 2006. “Cities Deliver Broadband for less,” Wired News, March 7, 2003, available at www.wired.com /news/ business/0,1367,57927,00.html. Clarke, richard, Kevin Hassett, Zoya Ivanova, and lau -rence Kotlikoff, “Assessing the Gains from Telecom Com- petition,” nBE r Working Paper #10482 , Cambridge, Massachusetts, May 2004. Clay, Karen, ramayya Krishnan, Eric Wolff, and Danny Fernandes, “ retail strategies on the Web: Price and n on-Price Competition in the online Book Industry,” Journal of Industrial Economics , Vol. 50, no. 3, 2002 , pp. 351–367. Columbia Telecommunications Corporation, “Fiber o ptics for Government and Public Broadband: A Feasibil- ity study,” prepared for the City and County of san Fran-cisco, Columbia, Maryland, 2007. “Companies That Fought Cities on Wi-Fi, now rush to Join In,” Wall Street Journal, March 20, 2006. Corning, Inc.,“Bridging the last Mile: Access network Wireline Architectures,” White Paper W P 6300, Corning, new y ork, 2001. Corning, Inc., “Broadband Technology overview,” White Paper W P 6321, Corning, n ew york, 2005. Criterion Economics, “The Effects of ubiquitous Broad - band Adoption on Investment, Jobs, and the u.s . Econ-omy,” Washington, D.C., 2003. DiMaggio, Paul, et al., “From unequal Access to Differen - tiated use,” in Kathryn neckerman, ed., Social Inequality, r ussell s age Foundation, n ew york, n ew york, 2004. Downes, Tom, and shane Greenstein, “ understanding Why universal service obligations May Be unnecessary: The Private Development of local Internet Access Mar -kets,” unpublished draft, 2006. California Economic Policy Broadband for All? 20 P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A P uBl I C P o l I Cy I n s T I TuT E oF C AlI F o r n I A 21 “s.F. Picks Google Wi-Fi Team,” San Francisco Chronicle, April 6, 2006. “some Worries as san Francisco Goes Wireless,” New York Times, April 10, 2006. “Verizon’s Fios services Build Momentun,” C nET, August 1, 2006, available at http://news.com.com / Verizons+Fios+services+build+momentum /2100 -1034_ 3- 6101038.html. “What If They Built an urban Wireless network and Hardly Anyone used It?” New York Times, June 26, 2006. “Wi-Pie in the s ky,” The Economist, March 9, 2006. “With a Dish, Broadband Goes rural,” New York Times, n ovember 14, 2006. Morton, Fiona scott, Florian Zettelmeyer, and Jorge silva- r isso, “Consumer Information and Discrimination: Does the Internet Affect the Pricing of new Cars to Women and Minorities?” Quantitative Marketing and Economics, Vol. 1, 2003, pp. 65–92. Prieger, James, “The supply side of the Digital Divide: Is There Equal Availability in the Broadband Internet Access Market?” Economic Inquiry , Vol. 41, no. 2 , April 2003, pp. 346 –363. Prince, Jeff, “Measuring the Digital Divide: structural Estimation of Demand for Personal Computers,” unpub-lished manuscript, 2003. “Public, Private Collaboration To Design silicon Valley Wide Wireless network,” Joint Venture silicon Valley press release, January 26, s an Jose, California, 2006. About the Author Jed Kolko is a research fellow at the Public Policy Institute of California. Contributors Davin Reed provided research assistance for this edition of California Economic Policy. I would like to thank Richard Greene for editorial assistance and Dean Bonner and Mark Baldas-sare for PPIC sur vey data. I am grateful to Forrester Research for permission to use the Techno-graphics Benchmark data. Anne Neville (State of California Business, Transpor tation, and Housing Agency), Shane Greenstein (Nor thwestern Universit y), Dan Mazmanian (Universit y of Southern California and PPIC), Jaime Calleja Alderete (PPIC), Ellen Hanak (PPIC), Deborah Reed (PPIC), and Mark Baldassare (PPIC) provided detailed and thought ful comments on draf t versions of this repor t. Brian Moura ( Wireless Silicon Valley Initiative) provided background insights on municipal wireless initiatives. Par ticipants at the Telecommunications Policy Research Conference and at PPIC seminars of fered helpful suggestions. Responsibilit y for any errors is solely mine. Walter B. Hewlet tDirectorCenter for Computer Assisted Research in the Humanities Ki Suh ParkDesign and Managing Par tnerGruen Associates Constance L . Rice Co-DirectorThe Advancement Project Raymond L . WatsonVice Chairman of the Board EmeritusThe Ir vine Company Carol WhitesidePresidentGreat Valley Center Board of Directors Thomas C. Sut ton, ChairRetired Chairman and Chief E xecutive Of ficer Pacific Life Insurance Company Mark BaldassarePresident and Chief E xecutive Of ficerPublic Policy Institute of California Linda GriegoPresident and Chief E xecutive Of ficerGriego Enterprises, Inc. Edward K. HamiltonChairmanHamilton, Rabinovit z & Alschuler, Inc. Gary K. Har tFormer State Senator and Secretar y of EducationState of California The Public Policy Institute of California is a private, nonprofit research organization established in 19 9 4 with an endowment from William R. Hewlet t . The Institute conducts independent, objective, nonpartisan research on the economic, social, and political issues af fecting Californians. The Institute’s goal is to raise public awareness of these issues and give elected representatives and other public of ficials in California a more informed basis for developing policies and programs. Copyright © 20 07 by Public Policy Institute of CaliforniaAll rights reser vedSan Francisco, CA Shor t sections of tex t , not to exceed three paragraphs, may be quoted without writ ten permission provided that full at tribution is given to the source and the above copyright notice is included. PPIC does not take or suppor t positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, suppor t , or oppose any political par ties or candidates for public of fice. Research publications reflect the views of the authors and do not necessarily reflect the views of the staf f, of ficers, or Board of Directors of the Public Policy Institute of California. PUBLIC POLICY INS TIT U TE OF CALIFORNIA5 0 0 Washington Street , Suite 8 0 0 San Francisco, California 9 4111Telephone: (415) 2 91- 4 4 0 0 Fax: (415) 2 91- 4 4 01 w w w.ppic.org ISSN #1553-8737 California Economic Policy Broadband for All? PuBl IC Pol ICy Ins TITuTE o F CAlIF orn IA 500 Washington s treet, suite 800 s an Francisco, California 94111NON-PROFIT ORG.U.S. POSTAGE PAIDBRISBANE, CAPERMIT #83 In This Issue of CEP Who has broadband? Who can get it? are available free of charge on PPIC’s website www.ppic.org R E C E N T I S S U E S O F California Economic Policy Pay-or-Play Health Insurance Mandates: Lessons from California Lawns and Water Demand in California Trade with Mexico and California Jobs Are Businesses Fleeing the State? Interstate Business Relocation and Employment Change in California A Decade of Living Wages: What Have We Learned?" ["post_date_gmt"]=> string(19) "2017-05-20 09:39:06" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(10) "ep_707jkep" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:39:06" ["post_modified_gmt"]=> string(19) "2017-05-20 09:39:06" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(52) "http://148.62.4.17/wp-content/uploads/EP_707JKEP.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }