Key Takeaways

While most Californians have access to broadband, at least two million households (15%) still do not—a gap known as the digital divide. In 2021, California invested $6 billion through Senate Bill (SB) 156 to expand broadband infrastructure, address affordability, and promote digital literacy. Local entities across the state are in the midst of planning efforts. In this report, we present findings from the first year of implementation, drawing on statewide broadband data and interviews with 41 community partners, spread across 54 of California’s 58 counties. We find that:

• Low-income, rural, and tribal communities are on the wrong side of the digital divide. Challenges include geological barriers, low population density, high poverty, and limited ability to fund broadband projects. Migrant and mobile-home communities also often live in underserved pockets, even within regions that are generally well served. To support successful and equitable implementation of SB 156, we recommend that the state prioritize the needs of these underserved communities as it develops its digital equity plan and five-year action plan.→
• Local communities are pursuing funding from federal, state, local, and private sources. SB 156 has opened many local communities up to the idea of a publicly owned network, which has the benefit of reaching populations and areas that are currently unserved or underserved by private internet service providers. Greater outreach and technical assistance from the state could help support local entities in applying for the available public grant programs.→
• Natural disasters pose serious threats to broadband networks. More than 2.7 million Californians live in parts of the state that are at very high risk of wildfire. Many communities fluctuate from served to unserved when natural disasters or other causes of network damage occur. New infrastructure should include resiliency plans to maximize coverage during natural disasters. Building so-called “redundancy” into a network may increase costs significantly, but a full examination of costs is outside the scope of this report. However, making this investment would provide crucial reliability in the event of emergencies—a time when connectivity is badly needed.→
• State and federal maps overstate how many households have access to broadband. Developing more accurate and granular maps will be necessary to identify and expand access to underserved neighborhoods. The education sector could play an important role in collecting data on real-world broadband access and speed.→
• Affordability and digital literacy are major barriers to broadband adoption. Broadband costs $68 per month, on average, in the US. The Affordable Connectivity Program (ACP) provides $30 per month to qualified households, but participation is very low, with only 1.8 million California households enrolled (30% of eligible households) as of March 2023. Further, giving people the tools and resources they need to become digitally literate is an important step to encourage adoption. The state can mobilize its vast networks of adult education centers, public libraries, and community colleges to offer digital literacy training and promote participation in the ACP.→

The COVID-19 pandemic spotlighted the importance of universal broadband. Students needed reliable internet to receive instruction during school closures—and have an ongoing need to receive, complete, and turn in assignments using the digital platforms adopted during remote learning (Gao, Lafortune, and Hill 2020). Workers relied on the internet to telecommute and receive workforce training (Galperin et al. 2021); patients needed the internet to access telehealth (O’Shea at al. 2022); and the general public depended on the internet for social interaction and civic engagement (McClain et al. 2021).

Yet two million California households (15%) did not have broadband in 2020 (Hayes, Gao, and Starr 2022). Historically, the US has depended largely on private investments by telecommunications carriers and service providers to build, operate, and maintain broadband networks. But networks can be expensive to build and maintain in places where there may not be sufficient customers. Key barriers to universal broadband access include availability, affordability, and adoption (Figure 1). Lack of infrastructure availability has an outsized impact on rural areas, multi-unit dwellings in urban areas, and tribal communities. Meanwhile, lack of affordability and low levels of the digital literacy needed to adopt broadband disproportionately affect low-income communities, communities of color, and non-English-speaking communities.

To remove these barriers to universal broadband access, federal and state policymakers have enacted a number of landmark policies in recent years, spurring the largest broadband investments in history (Figure 2).

In August 2020, Governor Newsom signed an executive order (N-73-20) to improve digital connectivity across the state. As a result, the state developed a Broadband Action Plan in pursuit of three goals:

• Goal 1: All Californians have high-performance broadband available at home, schools, libraries, and businesses.
• Goal 2: All Californians have access to affordable broadband and the devices necessary to access the internet.
• Goal 3: All Californians can access training and support to enable digital inclusion.

Also in 2020, the FCC established the Rural Digital Opportunity Fund, which will disburse up to $20 billion over 10 years to bring high-speed fixed broadband services to rural households and small businesses.

Then in July 2021, the state passed Senate Bill (SB) 156, providing an unprecedented $6 billion multi-year investment to increase equitable, affordable access to high-speed internet across California. At the federal level, the Infrastructure Investment and Jobs Act (IIJA) of 2021 provided $65 billion toward broadband deployment and adoption nationwide. The largest program funded by the IIJA, the federal Broadband Equity, Access, and Deployment (BEAD) program, provided $42 billion to support funding planning, infrastructure deployment, and adoption programs in all 50 states. Some $3.5 billion of this BEAD funding is estimated to go to California.

The goal of this report is to provide a detailed description of the current landscape of broadband connectivity across the state in the context of the first year of implementation of SB 156—the state’s largest public investment in broadband infrastructure. For this analysis, we rely on 41 interviews with a diverse group of community partners—including chief technology officers at county offices of education, city and county governments, the California Public Utilities Commission’s regional broadband consortia, tribal nations and associations, and nonprofit organizations and associations. The interviews addressed a variety of broadband projects (related to infrastructure, affordability, and digital literacy) across 54 of California’s 58 counties, which collectively serve 92 percent of the state’s diverse population. Please refer to Technical Appendix A for details about the coverage.

Our report begins by providing an overview of the investments in public broadband infrastructure made through SB 165. We then describe the different approaches that local communities are taking to building broadband, including the range of public and private involvement. Next, we examine the communities that are most likely to lack reliable broadband access and the barriers they face. We then summarize some of the hopes and concerns related to SB 165 expressed by local entities and discuss the need to continue efforts to improve broadband affordability and adoption. We conclude with recommendations based on our research findings.

At the time of its passing, SB 156 was the largest public broadband infrastructure investment in the country. As shown in Figure 3, broadband infrastructure includes three different components (California Department of Technology 2022a):

• The “backbone” or “long haul” is the set of global and national internet networks.
• The “middle mile” is the physical mid-section of the infrastructure required to connect the global/national and local networks, thereby enabling internet connectivity for homes, business, and community institutions.
• The “last mile” is the final leg of the network connecting homes, businesses, and community institutions with local networks.

SB 165 focuses especially on supporting “middle mile” and “last mile” connections. Specifically, the bill included:

• $3.25 billion to build, operate, and maintain an open-access, state-owned middle-mile network (i.e., the physical fiber optic infrastructure)
• $2 billion to set up last-mile connections (i.e., connecting homes and businesses with local networks)
• $750 million for a loan loss reserve fund to help local entities secure financing for broadband infrastructure

The 2022–23 state budget provided an additional $550 million toward the middle-mile project as it moves into construction. Because most of SB 156 ($5 billion) comes from the federal American Rescue Plan Act, the money must be encumbered (e.g., contracts awarded) by December 2024 and spent (i.e., construction completed) by December 2026.

Currently, middle-mile networks are largely owned by private internet service providers, which have not built out sufficient services to rural communities and low-income communities due to high costs. SB 156 would create an open-access network, which would give all providers and entities access to broadband infrastructure and has the potential to increase competition and drive down costs. The design and construction of the middle-mile network is monitored by the Middle-Mile Advisory Committee; Caltrans is conducting the middle-mile preconstruction work.

The California Public Utilities Commission (CPUC) recently released its Federal Funding Account Priority Areas map for last-mile funding, and is in the process of updating the map based on public input and feedback. The California Department of Technology (CDT) has also released a Statewide Construction Evaluation Map, representing 10,000 miles of proposed builds that span the state (Figure 4). It was based on the map proposed by GoldenStateNet—the state’s middle-mile third-party administrator—and prioritizes areas that can be built expeditiously, areas with no middle-mile access, last-mile residences that are unserved (defined as having less than 25 megabits per second, or mbps, download and 3 mbps upload service), and regions without sufficient middle-mile infrastructure to meet future needs. Because it may be cost-prohibitive to build all 10,000 miles, CDT is evaluating affordability and identifying segments that may be built or leased using Indefeasible Right of Use (IRUs).

The $2 billion going to last-mile connections will be issued through the Federal Funding Account and administered by CPUC. Eligible areas are those unserved by a wireline provider that can reliably provide 25/3 mbps service.

Another CPUC program, the $50 million Local Agency Technical Assistance (LATA) grant, is available to reimburse local governments and tribal entities for preconstruction costs, such as consultant services, environmental permitting, and engineering and design activities, as well as costs to form a broadband joint power authority. Applications for the LATA program opened in summer 2022, and as of March 2023, the total requests from non-tribal communities have exceeded the available funding ($45 million). The Broadband Loan Loss Reserve Fund provides $750 million to help local governments, tribes, and nonprofits secure financing for building last-mile projects. CPUC released a staff proposal for implementing this fund in September 2022.

As the state makes inroads on SB 156 implementation, local communities across the state are pursuing a variety of broadband projects. In this section, we explore various approaches to building out broadband networks and the range of public and private involvement.

Increased federal and state infrastructure funds have enabled local communities—including counties, cities, education agencies, and joint power authorities and organizations—to further their broadband efforts on the ground. While many communities are at the beginning of planning broadband projects, others have had successful projects for years. For example, the Digital 395 project, along Highway 395 in the Eastern Sierras, was completed decades ago and went into service during the winter of 2013–14. Meanwhile, BorderLink in Imperial County, which launched its pilot program in 2017–18, is an example of a project that was completed more recently.

However, most projects are still nascent and local communities are pursuing various models: many are negotiating with private companies for workable partnerships on the infrastructure build and maintenance, but the advent of SB 156 has opened many communities up to the idea of a publicly owned network. A community’s own network can take multiple forms, from a full municipal broadband to a full private network (Figure 5).

Figure 5 is adapted from a widely used model to characterize the various elements of broadband projects, including infrastructure ownership, operation and maintenance, and broadband delivery (US Ignite 2020). A full municipal broadband allows cities and municipalities to build and operate the network themselves, while a full private broadband is better able to attract private partners. Although most of the local projects are still in the early stages (i.e., without an engineering design or cost estimate), many of our interviewees indicated a preference for more public involvement—citing the failure of existing internet service providers (ISPs) to serve communities that have historically faced barriers to access in other sectors.

Our interview respondents noted with considerable frequency that costs, partnerships, existing infrastructure, resources, voter preferences, and long-term fiscal outlook were major factors shaping local broadband projects. Please refer to Technical Appendices A and B for more details. A full municipal broadband network has the benefit of reaching populations and areas that are currently unserved or underserved by private ISPs; however, it requires significant financial commitment and political buy-in. On the other end of the spectrum, a full private broadband allows local entities to capitalize on existing infrastructure and can attract more private funding, but it may be hard to serve areas where the return on investment is not favorable.

Communities in rural areas report more difficulty convincing ISPs and other private organizations that additional builds are “worth the effort” and will turn a profit. By and large, local governments do not traditionally play a role in connecting their residents to wireless or satellite services; however, in places where essential infrastructure does not currently exist, public entities have had to take a bigger role and they sometimes rely on these alternative approaches (e.g., the publicly owned options outlined in Figure 5) to connect residents.

Rural communities are also more likely to report topographic challenges as reasons why ISPs have not conducted builds: high-risk wildfire areas, mountainous regions and high altitude, windy conditions, lack of roads, extreme temperatures, and even local soil quality affect whether infrastructure builds can be conducted in the first place and/or whether they can be maintained effectively.

Regional and topographic differences may also affect individual households’ ability to secure broadband connections: Extreme heat in the Coachella Valley continues to raise consumers’ electricity bills by requiring extra air conditioner use. The higher electricity bills strain consumers financially, leaving less room in their budgets to afford broadband. In the city of Chico, incumbent ISPs have not rebuilt infrastructure affected by recurring wildfires, as ISPs have found it uneconomical to build back infrastructure or to “underground” their services (which would increase resilience in case of future fires). New residents, as a result, have often foregone broadband connections entirely, citing exorbitant connection fees and other financial concerns.

While most of the broadband projects are still in the planning stages and subject to change, below we review public documents to describe three very different models of building broadband.

Full Municipal Broadband in Santa Clara County

Despite being home to Silicon Valley, Santa Clara has a persistent digital divide that excludes many low-income families from broadband access. Nearly 100,000 residents in San Jose lack broadband access, including 36 percent of Latino families and 47 percent of Black families (San Jose Digital Inclusion Fund 2022). In December 2021, the Santa Clara County Board of Supervisors approved creating an ISP operated solely by the county or with other local jurisdictions to expand internet access to residents who lack reliable internet access or only have access through one provider. It is modeled after Chattanooga, Tennessee, a city that is well known for being the first community with citywide access to fast residential internet connections.

As of January 2023, Santa Clara County has an agreement with a consultant to prepare a broadband master plan, and is considering options for how to deploy high-speed internet, including fiber optic networks.

Public-Private Partnership in Los Angeles County

Los Angeles County has identified digital divide target areas in which more than 20 percent of the households lack internet subscriptions, according to estimates from the US Census Bureau’s American Community Survey. The vast majority of these estimated 416,000 households are composed of people of color and/or low-income individuals. In November 2021, the Los Angeles County Board of Supervisors approved a motion to invest in immediate- and longer-term efforts to facilitate affordable and reliable internet services to underserved communities. The county is not equipped to carry out the logistics to directly provide internet service in underserved and low-income communities due to a variety of constraints related to funding, resources, and technical capacity. Therefore, it administers a competitive solicitation process to contract with one or more managed service providers to implement and manage its Community Broadband Network pilot projects.

To date, 12 companies have responded to the county’s Request for Statement of Qualifications, which remains open until September 2027. These proposals utilize different fixed wireless technologies as well as fiber optic solutions. Some proposals include cost models that may offer economies of scale to deliver broadband to multiple communities at lower costs than initially projected. The county recently announced three pilot projects in late 2022 that will cover underserved communities in which more than 20 percent of households lack internet services: East Los Angeles/Boyle Heights, South Los Angeles, and the San Fernando Valley. The county released a competitive work order solicitation in January 2023 to deploy the Community Broadband Network in East Los Angeles/Boyle Heights and South Los Angeles.

The county anticipates that funding for these projects will come from a mix of federal, state, and local sources, including up to $56 million in the county’s American Rescue Plan Act (ARPA) State and Local Fiscal Recovery Funds allocation.

Open-Access Municipal Broadband Fiber Network in Alpine and Monterey Counties

The Rural County Representatives of California (RCRC), a 39-member county service organization representing rural counties, initiated the formation of Golden State Connect Authority, a joint powers authority representing the same 39 member counties, to install fiber using an open-access municipal broadband network. The network is modeled after UTOPIA Fiber, which was built in 2004 in Utah.

An open-access municipal network allows multiple ISPs to use the same publicly owned network to offer competing services to homes and businesses. For example, UTOPIA Fiber has 14 residential ISPs on its network. This spurs competition among ISPs and can drive down costs and improve quality.

In summer 2022, RCRC partnered with the South Salinas Valley Broadband Authority and Alpine County Board of Supervisors to install internet infrastructure in Salinas Valley along the Central Coast and in Alpine County on the Nevada border just south of Lake Tahoe. RCRC is conducting network planning and identifying underserved residents and businesses in those two counties. RCRC plans to expand to additional areas within rural California.

In this section, we consider whether there is a link between poor internet connectivity and the demographics of the community. We also take a deep dive into the barriers faced by different communities in trying to achieve reliable broadband access.

First, we examine the relationship between the median maximum advertised download speeds at the census tract level, and the demographic characteristics of that census tract: the percent of the population at or below 150 percent of the federal poverty level, percent of the population that is Black, percent of the population that is Latino, percent of the population that is Native American, and population density.

The results show that lower socioeconomic status and a larger share of Native Americans in the population are strongly associated with the digital divide. A single percentage point increase in the share of low-income households is associated with a roughly 1 mbps decrease in internet speed. A single percentage point increase in the share of Native Americans in the population is associated with a nearly 4 mbps decrease (Technical Appendix Table B3).

Our interviews, presented in detail below, provide rich context for better understanding barriers to internet access in rural, low-income, tribal, and itinerant communities.

Rural Communities Contend with Geographic Barriers

Rural areas face both geologic and topographical barriers to gaining internet access, according to the experts whom we interviewed. Topographically, interviewees in areas with a great deal of variation in high and low elevation levels often mentioned that even households that live next to, but at a slightly lower elevation than, households with broadband access may completely lack access themselves because they are not in the line of sight for radio waves or other wireless signal transmissions coming from a local peak. Trees can also block line-of-sight transmissions, often affecting these same areas with topographical challenges. To make the situation more challenging, wireless coverage networks vary across different companies.

Another common terrain challenge that has left many mountainous, inland areas of California underserved is the hardness of the rock that would need to be drilled to put fiber lines underground. Furthermore, our interviewees mentioned that difficulties in getting vehicles and equipment up to these mountainous or otherwise difficult-to-reach areas to maintain the broadband network is another reason rural communities are often underserved. This maintenance challenge is intensified by the fact that these areas are often far from the nearest city in which maintenance workers from ISPs are stationed.

Low-Income Urban Communities See Digital Discrimination

In urban settings, network lines from nearby last-mile broadband networks often do not reach into low-income neighborhoods. These disparities are particularly prominent in multi-unit dwellings. Not having broadband accessible in their housing unit limits residents’ ability to make use of subsidies like the Affordable Connectivity Program, which pays up to $30 of a household’s monthly internet bill; use of these subsidies might make delivery of service profitable to ISPs. Even those living in federally subsidized housing through the Section 8 program for low-income families are sometimes underserved, according to our interviewees. This lack of access is sometimes known as digital redlining—that is, the exclusion of certain demographic groups, low-income individuals in this case, from access to digital resources. This lack of access is often attributed to the low profits that ISPs expect to make if they expand service to these areas. This type of digital discrimination is not illegal as there are no regulations that dictate where to build out services.

However, in February 2022, the FCC formed a cross-agency task force to create rules and policies “to promote equal access to broadband across the country, regardless of zip code, income level, ethnicity, race, religion, or national origin” (FCC 2022). The FCC released its Preventing Digital Discrimination Notice of Proposed Rulemaking in November 2022.

Tribal Communities Face an Array of Obstacles

Households in tribal areas are less likely to have broadband—in part due to rugged terrain, low population density, and low income levels, which make these areas less attractive to private ISPs. However, tribal communities also face additional obstacles.

First, US tax code restricts tribes’ ability to issue tax-free government bonds, severely limiting financing options to fund broadband projects (Gregg 2021). As a result, many tribes that are entirely unserved by both ISPs and electricity providers rely exclusively on federal funding programs to build out infrastructure projects.

Second, many tribes lack the resources, staff, and capacity to prepare and submit a successful grant application. One respondent noted that he had been a telecom consultant and manager, “so I have some grounding in this that allows me to at least tread water,” but that, for the rest of his staff, “the flood of information about broadband is a challenge in and of itself.”

Third, poor data quality about the Native American population can result in fewer resources allocated to tribal communities than are needed. The FCC and the National Telecommunications and Information Administration (NTIA) have a number of programs that will provide funding for broadband deployment on tribal lands, but widespread undercounting of tribal communities—for example, Native Americans are often misclassified as white unless they are formally enrolled in a tribe—may impact federal funding, leaving these communities further behind (Diebold 2022).

Migrant and Mobile-Home Communities Are often Overlooked

Roughly 80,000 migrant youth live in California—one of every three migrant students nationwide. This underserved population is particularly in need of internet access (California Department of Education 2021). However, interviews with local leaders in one of the most populous counties in the state reveal that local migrant camps are usually severely underserved due to lack of infrastructure and/or affordability. Interviewees pointed out that itinerant workers have less incentive to pay for service to a residence that they occupy for only part of the year—especially since they have no guarantee of being able to return the following year.

Similarly, mobile-home parks are often “doughnut holes” in their broader communities’ networks—meaning that they represent small patches that lack service, even in communities that are generally well served, such as Santa Clara County. According to the interviewees, the shallow builds of the infrastructure for mobile-home parks are sometimes referred to as “spaghetti” because they are a chaotic mix of utility connections that sit near the street’s surface; this state of affairs creates challenges for adding broadband underground in those areas and can even pose physical dangers for potential builders.

The persistent digital divide highlights the limits of privately owned broadband networks. SB 156 aims to overcome these limitations. Major SB 156 implementation activities are currently underway across the state. At the federal level, the newly passed Infrastructure Investment and Jobs Act provided additional funding to support state, local, and regional efforts. At the state level, CalTrans is doing preconstruction work on the proposed statewide middle-mile network, and CPUC is working on rules and policies regarding last-mile grant programs. At the local level, communities are gearing up to take advantage of SB 156.

In this section, we summarize some of the hopes and concerns that our interviewees expressed about a successful and equitable implementation of SB 156.

Local Communities Look to a Variety of Funding Sources

All community partners applauded the historic investment of SB 156, which they believe will spur improvements in broadband infrastructure across the state. As one community leader noted, SB 156 will not immediately fix everything, but it will encourage “a state, county, and regional conversation that will act as a catalyst to close the digital divide.”

At the same time, given the vast diversity in geography and topography across the state, SB 156 may fall short. Several local entities mentioned that it costs about $65,000 per mile to install fiber in flat areas, but costs could increase to $500,000 per mile in mountainous terrains. Rising costs due to labor shortages, global supply chain issues, and inflation add additional pressures.

The passage of the federal Infrastructure Investment and Jobs Act provided much-needed additional support. As a result, local communities are actively looking into a combination of funding opportunities, including federal (e.g., NTIA, Capital Projects Fund, USDA), state (e.g., SB 156 Last Mile), local, and private funding sources.

Some communities are also exploring the possibility of using general funds, issuing general obligation bonds, or leveraging private philanthropic support. One challenge, however, is that not all local entities know how to coordinate these funds and where to prioritize, given that these programs serve different purposes and have different deadlines. For example, certain federal grant requirements prevent some tribes from accessing and maximizing critical funding, and the administrative burden can be overwhelming for many tribes (Office of Management and Budget 2022). Nonprofit organizations, including associations of governments, sometimes provide webinars, training, and resources to help local governments and tribes understand the eligibility criteria for each funding program.

Funding Processes often Disadvantage Small, Rural, and Tribal Communities

Applications for the Federal Funding Account (FFA) Last Mile program and the Loan Loss Reserve program are not yet being accepted. But representatives whom we interviewed from local communities raised several issues about the application process, citing their past and ongoing experiences with the California Advanced Services Fund (CASF) and Local Agency Technical Assistance (LATA) programs.

First, awareness of these programs is not uniform across the state, and many local communities are not knowledgeable about program goals, eligibility, deadlines, and application processes. For example, one large county in southern California cited very low awareness of the LATA program among local government agencies and tribes ahead of the application deadline. Respondents noted the limitations of CPUC’s outreach, especially among remote and rural areas of the state.

Second, the application process is often too complicated and the window is too short, burdening small, rural and tribal entities that do not have the staff, resources, or capacity to quickly put together a successful grant application. As a result, some counties, anticipating the short timeline of the application and build process, have already started taking steps to be shovel-ready. However, this approach tends to favor counties with more resources that have the ability to plan in advance and partner with private organizations or incumbent ISPs.

Third, several government associations—such as associations of governments and other regional organizations—mentioned that they are well equipped to provide technical assistance to local agencies. They are also familiar with the work required for environmental studies, as associations of governments often conduct needs assessments and develop strategic plans. But these associations are not eligible to apply for and distribute LATA funding.

Permitting Remains a Formidable Challenge

Because most of the SB 156 money is from the American Rescue Plan Act (ARPA), these funds must be encumbered by the December 2024 federal deadline and spent (i.e., projects built) by December 2026. Most interviewees expressed concerns about meeting the tight federal deadlines, especially in light of the extensive review and permitting process.

A major broadband project will need to undergo a number of federal and state permitting processes—including those required by the California Environmental Quality Act (CEQA), the State Historic Preservation Office, the California Department of Fish and Wildlife, the State Water Resources Control Board/Regional Water Quality Boards, the State Fire Marshal, the National Environmental Policy Act, and the US Fish and Wildlife Services. This process typically takes up to 30 months (California Department of Technology 2023). SB 156 included statuary exemptions from CEQA for broadband deployment projects along the right-of-way of existing thoroughfares (provided they meet certain other requirements), reducing the average permitting time by 13 months. Even so, the remaining state and federal permitting can take up to 17 months.

Existing Data Overstate Service Areas

The FCC and CPUC measure broadband deployment based on ISP reports. But companies can report an entire local area (a census block) as served if they offer service to at least one home or business, which leads to overstating access. In addition, advertised speeds do not always hold up in real-world tests. In places as diverse as the state of Georgia and the city of Fresno, there is clear evidence that this method overstates the extent of broadband deployment.

Georgia employed a location-based methodology, which is different from FCC and CPUC’s census-block-based approach. The new methodology more precisely maps broadband availability to every location (i.e., homes and businesses) and the results are striking. The Georgia Broadband Map identified more than half a million locations that lacked broadband availability, and about half of those (255,067) had previously been identified as served by the FCC (Georgia Department of Community Affairs 2020).

Similar work by Fresno Unified School District belies CPUC’s map, which shows only 5,000 underserved students in the city, based on our conversation with the district. Fresno Unified’s own speed testing found that, for academic year 2022–23, around 15,000 students were experiencing internet at less than 25 mbps. The average household size for families with school-age children is four, based on the American Community Survey, and this means that as many as 60,000 residents within a city of 500,000 were unserved. Fresno Unified’s data also show that over 49,700 students were experiencing internet at less than the state’s recommended 100 mbps, with a median download speed of 48 mbps. The district conducted multiple speed tests per district site during the academic year and observed a much lower rate of service than advertised. Nearly 59 percent of observations showed the customer as underserved—and 10 percent of those qualified as unserved (Figure 6). For detailed maps showing the extent of unserved and underserved residents in Fresno, see Technical Appendix Figure B1.

Because major federal and state programs from the FCC, USDA, and CPUC allocate funding based on the FCC’s maps, the inaccuracies oftentimes present a key barrier to identifying and reaching unserved and underserved populations. CDT’s middle-mile network is also based on CPUC’s broadband availability map—the inaccuracies in the data may mean that some unserved and underserved areas will be left out of the statewide network. For example, interviewees have told us that, historically, the overreporting of service availability in remote rural areas and dense urban areas, especially those with many multi-unit dwellings, has limited public investment in reaching those same communities, even as the private market has continued to find it unprofitable to serve them.

Recognizing the limits in the current data, the federal Broadband Deployment Accuracy and Technological Availability Act (2020) tasked the FCC with producing a location-based map that more precisely assesses where broadband is lacking. In November 2022, the FCC published its Broadband Serviceable Location Fabric (Fabric), which is a dataset of all locations in the United States where fixed broadband internet access is or could be installed. To generate the map, service providers’ availability data have been matched to the location information contained in the Fabric, rather than relying on a census block as the geographic unit. The FCC has created a process by which providers, states, and local and tribal governments can submit information to improve the accuracy of the map.

Natural Disasters Pose Serious Threats to Broadband Networks

For a broadband network, resilience is the ability to keep its users served even when one conduit of connectivity goes down. Natural disasters pose ongoing threats to broadband networks, underscoring the importance of centering resiliency planning in broadband deployment strategies. More than 2.7 million Californians live in parts of the state deemed at very high risk of wildfire, based on Cal Fire’s wildfire danger map. California has at least 75 communities in which at least 90 percent of residents live in very high-risk areas (Karlamangla 2021).

Without adequate resilience, many communities fluctuate from served to unserved in the face of natural disasters or other causes of network damage. Several regions in the state have had fiber and poles burn during fire season, leaving nearby neighborhoods disconnected for weeks at a time. Flooding, high winds, earthquakes, heavy snowfall, and ice have the potential to seriously damage both underground cables and those placed on poles. CPUC has highlighted significant problems in the state’s communications infrastructure in relation to natural disasters. For example, throttling (i.e., intentionally slowing down) communications networks undermined the Santa Clara County Fire Department’s ability to deploy relief efforts during the Mendocino Complex fire (CPUC 2020).

As natural disasters increase in both frequency and intensity, state and local governments as well as ISPs must take renewed steps to ensure the broadband network remains operational. Having redundant points—multiple entry points that can pick up the slack if one goes down—built into the network ensures everyday network stability. One challenge is funding, as developing resiliency plans, advancing technological solutions, and implementing these programs require a significant amount of financial investment, increasing costs in areas that are already underserved.

Labor Shortages and Supply Chain Issues May Cause Delays

COVID-19 has exacerbated labor shortages and strained the global supply chain, adding more pressures at a time when local communities are already facing a tight funding timeline. A recent survey from the Associated General Contractors of America found that 91 percent of construction firms have a hard time finding workers to hire, driving up costs and delaying projects across the country (Associated General Contractors of America 2022). On average, global container shipping rates have more than quadrupled since 2019, and schedule delays have risen exponentially. In some key trading routes, such as those from Asia to North America, the rate spikes are even higher and the delays more frequent (Bartman 2022). The state, through its competitive bidding process, secured the necessary conduits, vaults, optical fiber, and construction hardware for the next four years as it begins construction on the middle-mile network (Office of Governor Newsom 2022). Yet local communities that are leading various last-mile projects may face more uncertainties and project delays.

The expansion of broadband infrastructure—spurred by SB 156 and other major federal investments—could drive down costs and improve affordability in the long term. In this section, we examine barriers to affordability and adoption in the short term.

The average cost of broadband is $68 per month in the US, which is more expensive than in Europe, where the average cost is $45 per month (New America 2020). To help households offset the cost of broadband, the FCC launched the Emergency Broadband Benefit program in May 2021, providing $50 per month to qualified households (e.g., households with income below 200% of the federal poverty level or those receiving means-tested assistance programs). The program was replaced by the Affordable Connectivity Program (ACP) in 2022 to serve eligible households in the long term, but at a reduced benefit of $30 per month.

Few Californians Are Using the Affordable Connectivity Program

As of March 2023, participation is very low with only 1.8 million eligible households (30%) enrolled in the ACP in California (California Department of Technology 2023). The good news is that counties with higher need, such as counties with more low-income and unconnected households, have higher take-up rates (first two buttons in Figure 7).

Several factors have contributed to the overall low take-up rates:

• First, awareness is very low: nationwide, only 23 percent of adults have heard “a lot” or “some” about the Emergency Broadband Benefit—the predecessor of ACP (Sabin 2021).
• Second, many households may be discouraged by the multi-step application process, which requires them to first apply for the program and then arrange for a participating ISP to apply the subsidy. The website (affordableconnectivity.gov) may also be difficult to navigate for families, especially those without internet and those who do not speak English.
• Third, some ISPs have required enrollees to upgrade their existing data plans to more expensive plans, or to accept slower connection speeds if they want to apply those subsidies (Romm 2022).
• Fourth, in areas with high costs of living, such as the coastal regions of California, the subsidy amount of $30 is not sufficient to cover the full cost of getting connected, and families “have to choose between putting food on the table and paying for internet,” according to one interviewee. Accordingly, we find that counties with higher costs of living have lower ACP participation (third button in Figure 7).

Nearly all local entities interviewed are involved in efforts to promote ACP. Schools—a trusted messenger in local communities—have played an especially important role advertising ACP and encouraging eligible households to sign up. In 2022, the California Department of Education designated August 27 as the first Get Connected! Day, and schools across the state hosted enrollment events to help eligible families apply.

Improving Digital Literacy—the Last Barrier to Broadband

Even where problems of infrastructure and affordability have been solved, there remains the issue of digital literacy. People who do not have access to the internet will not necessarily be savvy about procuring it or using it.

Working parents, residents in underserved rural areas, and tribal communities may not have the capacity to wholly take advantage of full internet access. During distance learning, school districts noticed lower levels of participation from low-income parents and parents who do not speak English, as they often do not have the skills to help their children navigate digital resources or help with homework assignments. In health care settings, new immigrants who do not speak English also reported difficulty using digital tools to monitor blood pressure and communicate digitally with their health care providers (please see the text box “Widespread Consequences of Inequitable Broadband Access” for details and references).

Through our interviews, we found that educational institutions, including K–12 schools, county offices of education, community colleges, and public libraries, have played an important role in improving digital literacy and inclusion. Many school districts and county offices of education we spoke to provided informational flyers—Los Angeles and Santa Clara Counties offered information in Spanish and Tagalog, while the Imperial County Office of Education partnered with community-based organizations to distribute flyers. These districts and others hosted training programs, such as parents’ nights, to help students—and their family members—navigate new digital tools. Some counties, such as Sacramento, partnered with community colleges to recruit college students as “digital navigators,” and some cities, including San Jose, partnered with local libraries to provide computer classes and technical support. In addition, some regional consortia work with California State University campuses to focus on culturally competent and linguistically appropriate assistance. Our interviewees noted that these programs have achieved some level of success and provide good models to other communities across the state.

Broadband is critical for commercial, educational, and social purposes. While most Californians have access to broadband, at least two million households still do not. To help close this digital divide, federal and state programs have provided funding to support broadband deployment in unserved and underserved areas. SB 156 is the largest public broadband investment in California’s history. Major statewide implementation efforts are currently underway, and local communities are gearing up to expand broadband access in their areas. The first year of SB 156 implementation highlights reasons for hope—and some expected challenges—as the work continues.

To achieve universal broadband, we provide the following recommendations:

• The state needs to ensure that communities on the wrong side of the digital divide are included in the planning process. The state is currently developing its five-year action plan to capitalize on federal Broadband Equity, Access, and Deployment (BEAD) funding; it is also in the process of developing its digital equity plan. Both plans are required for the BEAD program and should aim to center the voices and prioritize the needs of low-income, rural, tribal, migrant, and mobile-home communities that are the most likely to lack broadband access.
• The state should improve current efforts to produce granular, accurate, and reliable broadband mapping. Mapping and data are key to identifying existing gaps, allocating investment dollars, and ensuring that the funding achieves the goal of reliable broadband for all. Yet detailed data on households lacking access to broadband remain limited. The FCC uses its broadband maps to allocate billions of dollars in subsidies, but it relies on data supplied by internet service providers. Companies can report that a census block is served even if only one household has internet service, which leads to overstating access. Using multiple data sources, the NTIA’s Indicators of Broadband Need revealed substantial gaps between what providers say they offer and the speeds consumers report using. At the state level, the California Public Utilities Commission’s broadband mapping has similar limitations. The FCC recently released its location-based data, which is an important step toward more granular, accurate, and reliable maps.
• The education sector can help improve broadband mapping. Nearly all districts have purchased devices for students, and schools could play an important role in collecting data on actual usage and speed. For example, the state can develop a speed-test application that schools may run on their devices.
• CPUC should expand its outreach and technical assistance efforts. Many factors make it challenging for small, rural, and tribal communities to prepare and submit a successful grant application. For example, not all eligible local entities are aware of all grant programs, the eligibility criteria vary from program to program, the application process can be complicated, and the application window can be short. CPUC could partner with county offices of education and school districts, which have already been spearheading local broadband expansion efforts, to raise awareness about all major programs. CPUC may also prioritize small, rural, and tribal communities in its outreach and technical assistance efforts so that these communities are on an equal footing during the application process.
• Develop resiliency plans when designing and building new infrastructure. More than 2.7 million Californians live in parts of the state that are at very high risk of wildfire, according to Cal Fire. As natural disasters increase in both frequency and intensity, state and local governments as well as internet service providers must take renewed steps to ensure that broadband networks remain operational. Though building in redundancy helps ensure network stability, it can also be financially prohibitive. Nevertheless, considering the resiliency of new networks should be a critical part of the state’s efforts to achieve universal broadband.
• Assess the need for additional subsidies to reflect the prevailing cost of living or the cost of delivering broadband to different areas. The average broadband cost is $68 per month in the US, which is higher than in most OECD countries (New America 2020). In coastal regions with high costs of living, the current subsidy ($30/month) may be insufficient. While the buildout of a statewide middle-mile network could spur competition, which could in turn drive down costs, in the short and medium term, the state should consider commissioning a study to examine how existing subsidy programs have been implemented and assess the need for additional subsidies.
• The state can leverage adult education centers, public libraries, and community colleges to provide digital literacy training. Many of these institutions have existing digital literacy programs, but they may not be well known to the larger community, and even to local leaders in charge of broadband expansion. The state should also consider building on the success of such outreach events as the first Get Connected! Day, sponsored by the California Department of Education. This event mobilized educational institutions, businesses, and community-based organizations to promote awareness and encourage participation in programs designed to expand digital access.
• Private philanthropy can support a statewide network of practitioners to share local practices and spotlight innovations. Currently, there is no database that collects information on local broadband projects. Among the local entities we interviewed, the vast majority expressed a strong desire to exchange information and learn from other local communities. Private philanthropies can establish a community of practice, convene local partners, facilitate the exchange of ideas, and shine a light on innovative programs that could be scaled across the state.

The COVID-19 pandemic reinforced the urgent need to build out broadband infrastructure and bridge the digital divide. Recent federal and state investments provide unique opportunities for local communities to plan, develop, expand, and sustain their broadband programs. Coordinated efforts among various stakeholders—including service providers, municipalities, public and private agencies, philanthropies, and community organizations—will be necessary to leverage federal and state funding, scale up effective innovations, and ultimately achieve universal broadband infrastructure.