BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA

(Filed August 6, 2021)

OPENING COMMENTS OF THE SAN DIEGO ASSOCIATION OF GOVERNMENTS (SANDAG) ON THE ASSIGNED COMMISSIONER’S RULING REGARDING BROADBAND INFRASTRUCTURE DEPLOYMENT

Kirk Blackburn Legislative & Regulatory Advocate Ellison Wilson Advocacy, LLC 1201 K Street, Suite 1201

Sacramento, CA 95814

Tel: (916) 448-2187

September 3, 2021                                          E-mail: lobby@ellisonwilson.com Advocate for SANDAG

BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA

Order Instituting Rulemaking Regarding Broadband Infrastructure Deployment and to Support Service Providers in the State of California

Rulemaking 20-09-001

(Filed August 6, 2021)

OPENING COMMENTS OF THE SAN DIEGO ASSOCIATION OF GOVERNMENTS (SANDAG) ON THE ASSIGNED COMMISSIONER’S RULING REGARDING BROADBAND INFRASTRUCTURE DEPLOYMENT

I.                            Introduction

The San Diego Association of Governments (SANDAG) submits these opening comments on the Assigned Commissioner’s Ruling Regarding Broadband Infrastructure Deployment and to Support Service Providers in the State of California (Rulemaking 20-09- 001). On August 20, 2021, the Administrative Law Judge issued a ruling extending the deadline to provide opening comments to September 3, 2021. SANDAG serves as the Regional Transportation Planning Agency and the Metropolitan Planning Organization for the San Diego region. SANDAG is governed by leaders of 19 local jurisdictions including 18 cities and the County of San Diego, representing more than 3 million residents.

Bridging the digital divide and ensuring equitable access to broadband is a priority for SANDAG. Access to broadband is essential for modern life and enables access to opportunities including education, jobs, sustainability, public health, and safety. Broadband is also critical to the future of transportation. As the regional transportation planning agency, one of SANDAG’s primary responsibilities is to develop a long-range transportation plan that serves as the blueprint for the transportation system. The Draft 2021 Regional Plan outlines a transformational vision for the San Diego region, harnessing advancements in technology to create a transportation

system that is faster, fairer, and cleaner. This visionary transportation system relies on digital infrastructure and broadband connectivity to integrate and orchestrate the transportation network, which is critical to meeting our goals.

In response to Governor Newsom’s Executive Order N-73-20 in August 2020, the SANDAG Board of Directors adopted a Board Resolution in January 2021 recognizing the critical nexus between broadband access and the region’s quality of life. The SANDAG Board is committed to developing a Digital Equity Strategy and Action Plan¹ (Strategy) that leads to rapid broadband deployment and adoption in the San Diego region and that supports the goals of the California Broadband Action Plan. To help inform the development of the Strategy, SANDAG formed

a Regional Digital Divide Taskforce comprised of representatives from public agencies, broadband providers, education, non-profit organizations, and other stakeholders actively working to bridge the digital divide. Concurrent to that effort, SANDAG began collaborating with a larger Southern California-wide initiative called Southern California Transformation to coordinate and align work efforts for Southern California’s residents. This group is inclusive of each county within the Southern California Association of Governments (SCAG) and SANDAG regions, Internet Service Providers (ISPs), County Supervisors, and other elected officials, in addition to the California Emerging Technology Fund and National CORE, an affordable housing developer throughout Southern California. SANDAG has also partnered with SCAG to develop consistent broadband-friendly policies and a coordinated approach to rapid broadband deployment across Southern California. Working together, SANDAG and SCAG can address the digital divide for approximately 60% of the state’s population.

In addition to developing the State Broadband Action Plan, the Governor’s Executive Order tasked many state agencies to take specific actions to expand high-speed broadband statewide. Caltrans is one of the agencies identified to develop a “Dig Smart” policy that identifies corridors that serve as strategic opportunities to integrate the infrastructure needed to expand broadband services. Leveraging existing or planned public infrastructure projects to expand broadband infrastructure is a key strategy for SANDAG and our member agencies to support the build-out of the network. As a council of governments tasked with planning, designing, and building transportation projects, SANDAG is well-positioned to coordinate the integration of fiber into transportation projects in areas that are unserved. SANDAG has been coordinating

¹ https://www.sandag.org/index.asp?classid=13&projectid=614&fuseaction=projects.detail

closely with Caltrans District 11 to pursue Dig Smart or Dig Once opportunities in the region starting with a demonstration project that will add fiber to an 18-mile roadway rehabilitation project along State Route 67 ² in San Diego County. Integrating conduit and fiber into the planned pavement work along the corridor that today lacks fiber infrastructure will enable the provision of broadband services to the 225,000 residents that live along the corridor. The SR 67 dig once demonstration is the first-of-its-kind in the state to demonstrate how public infrastructure projects can be coordinated with internet service providers (ISPs) to expand broadband service in areas that are hard to serve given geographic constraints or low demand. SANDAG is also coordinating with local jurisdictions, tribes, and utility provider San Diego Gas & Electric (SDG&E), to identify other key public works projects that could be leveraged to build out the network in areas where there are gaps. Lastly, SANDAG is in touch with the Southern Border Broadband Consortium to align work efforts and address middle mile gaps and infrastructure investments in both Imperial Valley and San Diego County, specifically along State Highway Routes 78 and 94-98.

II.                       Response to Comments

To support the expansion of broadband infrastructure in the region and state, SANDAG respectfully submits the following comments to the six issues identified in the assigned

commissioner’s ruling. Comments are consistent with SANDAG’s Strategy to expand broadband access and adoption in the San Diego region and include:

• Strategic and coordinated deployment: Corridors proposed by the CPUC’s “Anchor Build Fiber Highways” network omit key priority corridors in the San Diego region that could connect rural, tribal, and high-poverty underserved neighborhoods in urban and rural areas. Many regional and local governments have parallel broadband infrastructure planning and deployment efforts underway that should be integrated

into the CPUC’s proposed network. SANDAG urges CPUC to collaborate closely with state and regional stakeholders such as MPOs, tribal nations, Caltrans, local jurisdictions, and ISPs to refine the locations of the middle-mile network to prioritize public investment in areas of greatest need and complement last-mile deployments.

² https://www.sandag.org/index.asp?newsid=1291&fuseaction=news.detail

• Future-proof the network: Modern society’s reliance on information and communications technology is rapidly growing. SANDAG recommends that the CPUC deploy a resilient and integrated open access middle-mile network that can adequately meet future demands, beyond the State’s 100/20 Mbps broadband standards.
  • Data-driven decision-making: More granular, accurate data is necessary to inform the CPUC’s decisions for the middle-mile network. As the state’s regulatory agency, the CPUC is best position to acquire household-level and broadband infrastructure data from ISPs to inform decision-making and ensure that middle-mile network complements last mile deployments provide high-quality and affordable broadband to unserved and underserved areas.
• Identifying Existing Middle-Mile Infrastructure

What routes, if any, should be modified, removed from consideration, or revised?

Provide an explanation for these suggestions?

There are clear gaps in fiber infrastructure in the County of San Diego where public investment in the middle-mile network could fill. The CPUC should consider leveraging public infrastructure beyond the state highway system like County and tribal roads, transit, and rail right of ways, and public utility right of way projects to maximize the use of public funds in areas of greatest need. As proposed, the Commission’s “Anchor Build Fiber Highways” omits the state’s first dig once demonstration, SR 67 as well as several key corridors in San Diego County such as segments SR 78, SR 79, and SR 76 that could provide critical broadband connections to tribes and rural communities like the La Jolla Reservation, Los Coyotes Reservation and the communities of Warner Springs, Ranchita and Borrego Springs. As mentioned, SANDAG is currently working with Caltrans, the County of San Diego, and SDG&E to identify other local roads and utility projects that could be used to provide middle-mile connections. The CPUC should consider leveraging local roads in addition to the state highway system, such as S1, that could also serve as strategic connections to the network. Making the final determination of the middle- mile network locations without adequate consultation with key stakeholders like Caltrans, County Transportation Commissions, County Schools and Libraries, Metropolitan Planning Organizations, Local Jurisdictions, Tribal governments, as well as the California State Parks is premature and may not produce a network that serves the areas that are most impacted by the

digital divide. SANDAG urges the Commission to provide ample opportunity, beyond this rulemaking, for communities and stakeholders to provide feedback on the location of the open-access middle-mile network.

In addition, the middle-mile location identification process could be augmented with granular infrastructure data from ISPs to further refine the network locations. The CPUC should consider acquiring network location information from ISPs to design a middle-mile network that is open but not duplicative of existing networks where ISPs already provide affordable broadband service so public funding can be used to expand upon the broadband infrastructure that exists today. Ensuring that there are adequate connections to enable last-mile deployment in areas that are unserved and underserved is a top priority for SANDAG; acquiring granular network location and mapping information via the Commission is critical and could benefit public agencies in planning, permitting, and implementing more rapid broadband deployment.

Are there existing middle mile routes that are open access, with sufficient capacity, and at affordable rates on the county highway routes listed in Attachment A?

More granular data around service availability, quality, and affordability is lacking to make these determinations. An analysis of available data for San Diego County suggests that the end-users experience in terms of service quality, reliability, and affordability may be very different than what is shared and advertised to customers³. As mentioned earlier, SANDAG is working closely with Caltrans and other regional stakeholders to identify key corridors and routes that could serve as strategic opportunities to expand the middle-mile network. SANDAG urges the Commission to coordinate with agencies like SANDAG, Caltrans, County Transportation Commissions, Local Jurisdictions, Broadband Consortia, Tribal Governments, and ISPs to identify middle mile routes that may have sufficient capacity and may not be outlined in Attachment A. Employing a coordinated and strategic approach with other stakeholders will enable the Commission to leverage available infrastructure, coordinate construction to reduce cost, and allocate public funding to areas of greatest need.

In the context of these routes, what constitutes sufficient capacity and affordable rates?

In the context of these comments, SANDAG also recommends that the Commission develop middle-mile infrastructure that well exceeds federal broadband standards. Today, the Federal Communications Commission (FCC) defines broadband as internet service that meets

³ https://www.sandag.org/uploads/projectid/projectid_614_29513.pdf

download speeds of 25 Mbps and upload speeds of 3 Mbps, or 25/3 Mbps. These service standards are insufficient to complete essential daily tasks such as online learning, remote work, telehealth, especially when there are multiple devices in a household sharing bandwidth. While the state has established a new broadband threshold of 100 Mbps/20 Mbps, SANDAG encourages the CPUC to be forward-thinking and design a network that can meet current and future demand generated by the open-access middle-mile network. At a minimum, the network should be designed to provide 1 Gbps service to each household or 10 Gbps for larger services to businesses or public facilities like healthcare facilities, libraries, and schools. This is especially critical as many jurisdictions are undertaking “Smart City” initiatives that rely on technology to make municipal operations more efficient, sustainable, and in turn, allow cities to deliver digital services to residents. Additionally, a high-capacity and resilient network is critical for the future of transportation. An integrated network will enable transportation agencies like SANDAG and local jurisdictions to implement transportation technology solutions that support safety & mobility such as lane and detour management during major incidents, outages or evacuation management, real-time motorist information, smart intersections, real-time transit information, connected and autonomous vehicle communications, and curb or parking management.

Enabling the provision of high-quality, competitive, and affordable broadband service via the open-access middle-mile network is top of mind to ensure end users can realize the benefit of the network. Today, there are limited choices of providers in San Diego County. Most households in the region are served by at most two providers and rural communities and tribes have only one provider option. This lack of competition can lead to high costs for low-quality plans. In rural parts of the County, the most affordable plan is $90 per month4. Additionally, a Census analysis5 revealed that a high percentage of low-income households in communities across the County do not have broadband at home. Most of the communities are concentrated in the southern urbanized areas of the County and the unincorporated communities in East County, which collectively have a median household income of $35,000. These are the same communities that have the highest unemployment rates, face challenges with accessing transportation, education, and jobs. Providing affordable and competitive broadband offerings to

⁴ https://www.sandag.org/uploads/meetingid/meetingid_5885_29423.pdf

⁵ https://www.sandag.org/uploads/projectid/projectid_614_29513.pdf

these communities are paramount to their quality of life. SANDAG urges the CPUC to design network routes with end-user affordability and last-mile deployment in mind. Additionally, household broadband subscription data, including the cost of service, is not shared by ISPs.

SANDAG urges the Commission to acquire household-level data maintained by ISPs to inform the establishment of affordable service rates.

For routes that are identified as being open access, with sufficient capacity, and at affordable rates, how should the Commission verify these claims (e.g., should Communications Division send a data request for service term sheets, rates, approximate dark fiber, lit fiber, and conduit capacity, etc.)? Are there any other criteria that should be used to verify these claims?

Capacity and affordability claims must be handled in an accountable and transparent manner. SANDAG recommends the CPUC acquire granular mapping and data from ISPs to ensure claims are adequately verified and inform decision-making.

B.    Priority Areas

Is it reasonable to assume counties with a disproportionately high number of unserved households (e.g., 50% or more unserved at 100 Mbps download) are areas with insufficient middle-mile network access?

As the state’s regulatory agency, the CPUC is best positioned to acquire more robust, and transparent data from ISPs to inform decision-making. SANDAG recommends that the CPUC consider integrating public and private data from ISPs to ensure that the determination of unserved households is informed by accurate and reliable data.

San Diego County is a geographically diverse county and there are significant differences in broadband accessibility within the urban and rural areas. Data from FCC indicates that 94% of the population in the County’s urbanized area have access to fixed broadband that meets 25/3 Mbps, as compared to only 66% in the County’s rural areas.⁶ In order to accurately acknowledge and meet the different needs experienced by urban and rural communities within the County, SANDAG also recommends that the Commission utilize a more granular geography such as census blocks to determine areas with insufficient middle-mile network access. Identifying current unserved and underserved populations should, at minimum, utilize the state’s new broadband definition of 100/20 Mbps as opposed to 25/3 Mbps to ensure that populations have

⁶ https://www.sandag.org/uploads/projectid/projectid_614_29513.pdf

access to adequate broadband services that will meet their needs for essential online tasks. This should also include other factors that may impact broadband adoption and affordability such as future network demands, poverty rates, median household income, and other demographic factors to ensure that the number of unserved households is reflective of the level of need.

What other indicators, if any, should the Commission use to identify priority statewide open-access middle-mile broadband network locations (i.e., built expeditiously, areas with no known middle-mile network access, regions underserved by middle-mile networks, regions without sufficient capacity to meet future middle-mile needs)?

SANDAG recommends the CPUC integrate public and private data to inform decision-making. In addition to broadband subscription data, existing network locations, and speed test information from ISPs, this could also include public data such as population density, proximity to anchor institutions, or public facilities such as schools, hospitals, health clinics, public housing, and libraries. Other indicators that the CPUC may want to consider include communities with no middle mile network infrastructure and areas that lack adequate access to broadband service that meet the 100/20 Mbps thresholds, rural areas, tribal areas, areas with high poverty neighborhoods, senior populations (65+), at-risk youth/students, and areas with slow economic growth.

Many regional and local governments have parallel broadband infrastructure planning and deployment efforts underway that should be integrated to identify priority network locations such as the SR 67 dig once the project, utility undergrounding efforts, another prospective dig once corridors, or the FCC’s Rural Digital Opportunity Fund (RDOF). SANDAG encourages the CPUC to collaborate closely with state and regional stakeholders such as MPOs, tribes, Caltrans, local jurisdictions, and ISPs to refine the locations of the middle-mile network to prioritize public investment in areas of greatest need like rural communities and tribal areas.

C.   Assessing the Affordability of Middle Mile Infrastructure

What are existing providers paying or charging for middle-mile services?

The cost of leasing the middle mile from private middle-mile broadband providers is generally not available as published information. Some large wireless carriers lease middle-mile fiber from large fiber providers with nationwide agreements and rates that are not accessible to small last-mile providers in small communities. There are no published prices available from

private providers, but leasing rates from public providers⁷ indicate that public leasing offerings are significantly lower than those offered by private providers.

Are there other factors or sources of information the Commission should consider for determining whether these services are affordable?

In determining the affordability of the middle-mile network, the Commission should consider complementary investments in the middle-mile network that are being made by entities like SANDAG, local jurisdictions, transportation agencies, and other non-profit agencies like CENIC. With the significant amount of funding earmarked for broadband infrastructure from the state and federal government, SANDAG encourages the CPUC to work with state, regional, local, and private sector entities, to maximize the effective use of public funds to build out the open-access middle-mile network and ensure that broadband services provided to the consumer are affordable.

Is it reasonable for the costs of these services to change depending on the location where the service is provided (i.e., rural vs urban)?

In general, communities and tribes in the County’s rural areas lack broadband communications infrastructure, have limited access to broadband service (if at all), and experience disproportionately high costs. SANDAG concurs that it is reasonable for the cost of these services to change depending on the network location and strongly recommends that the CPUC take these factors into consideration when determining the costs of these services. The CPUC should coordinate with last-mile deployments to ensure that high-quality and affordable service options are fair, competitive and also provide affordable options for low-income households.

• Leasing Existing Infrastructure

If there is existing open-access communications infrastructure with sufficient capacity

to meet the state’s needs, should the state purchase IRUs from that network?

In planning and designing the location for the open-access middle-mile network, SANDAG recommends that the CPUC evaluate several business models or other alternative delivery models, that could support the rapid deployment of the network, prioritize areas of greatest need, and enable connections to last-mile deployment that meets minimum service quality and affordability standards. This could include Indefeasible Rights of Use (IRUs) or other public-private partnership (P3) models. SANDAG urges the Commission to consider

7 https://www.ctcnet.us/DarkFiberLease.pdf

evaluating these models using outcome-based performance measures or integrating concessions into these arrangements to ensure that network objective and end-user benefits are ultimately realized.

Is there any value in the state purchasing an IRU from the network if capacity is already available?

Consistent with our previous comment, SANDAG recommends that the Commission consider all available delivery models to determine the most effective approach for routes along with the middle-mile network. SANDAG urges the CPUC to acquire data from ISPs on existing network location information to determine the availability and capacity of the existing network. The Commission should also consider existing and future demands to ensure sufficient network redundancies are incorporated and public investment is maximized and prioritized in areas that are hardest to reach and serve.

If the state relies on IRUs for the development of the statewide network, will the generational investment that this funding provides be diminished when the IRU leases end 20 to 30 years later? Will existing networks run out of spare capacity?

At the pace that the technology revolution has grown, it is quite possible that broadband service of 100/20 Mbps, or 1 Gbps, may not be sufficient in several years. Network delivery models pursued by the CPUC could integrate outcome-based performance measures to ensure that network performance objectives are realized. This could also include terms that outline technology refresh cycles to ensure that service quality and capacity are satisfactorily being met, enable flexibility to scale or create redundancies in the network, and allow for the integration of new transmission technologies as the industry or demand changes. Designing a network for the future will ensure growing demands on broadband technology are met and provide cost savings by avoiding expensive retrenching in the future.

E.    Interconnection

At what points should the statewide network interconnect (e.g., to other networks, servers, etc.)?

Publicly-led fiber network deployments could serve as strategic interconnection points to establish a resilient and integrated statewide network. This could include networks such as the South Bay Fiber Network, which is a dedicated fiber-optic network developed for the use of

South Bay Cities Council of Governments (SBCCOG) and its member cities⁸ or the transportation fiber ring in San Diego County comprised of SANDAG, Caltrans, North County Transit District, and Metropolitan Transportation System. SANDAG recommends that the Commission continue to coordinate with public and private stakeholders such as Caltrans, County Transportation Commissions, Regional Metropolitan Organizations and Local Jurisdictions, broadband consortia, ISPs, and other stakeholders to identify interconnection points are adequately planned for.

Additionally, several public or private agencies may be interested in making connections to the middle-mile network in the future but may not have the resources or the expertise to conduct a detailed technical analysis to provide meaningful input on network interconnection points at this stage. SANDAG urges the Commission to provide ample opportunity, beyond this rulemaking, for stakeholders to provide feedback on interconnection points, especially to ensure last-mile deployments can be made and prioritized in areas of greatest needs.

Are additional exchange points necessary or strategic, and if so, where?

Consistent with our comments above, SANDAG recommends that the Commission coordinate with public and private stakeholders to identify additional strategic exchange points. These locations should prioritize connections in areas of greatest need while also making high-band redundant connections to ensure system reliability and future capacity to support 1 Gbps or 10 Gbps are achieved.

F.    Network Route Capacity

How many strands of fiber should the network deploy for each route?

The deployment of fiber along the middle-mile network should be based on a standard approach for estimating future demand beyond today’s 100/20 Mbps standard. The fiber strand count should be based on the total number of households and business services planned for the community being serviced and assume 1 Gbps service to each household or 10 Gbps for larger services to businesses. Additionally, consistent with the State’s Dig Smart-Dig Once policy framework, SANDAG encourages the CPUC to consider providing empty conduit duct banks with the intent of partnering with an ISP who may like to pull their own fiber. This approach could help spur market competition by enabling smaller ISPs to enter the market in a more cost-

⁸ https://www.southbaycities.org/programs/south-bay-fiber-network

effective approach. Based on our experience, SANDAG recommends that empty conduit banks be sized to account for the largest count fiber strand required to serve the communities and provide a minimum of four ducts of the same size, no smaller than the 1.5-inch inner diameter.

SANDAG also encourages that the Commission remains flexible as fiber needs may vary based on current and projected demand along the middle-mile network based on the number of end-users, applications, or other uses for communications infrastructure.

Are there other requirements or standards the Commission needs to consider to determine sufficient capacity?

As indicated earlier, SANDAG urges the Commission to develop a middle-mile network that meets today’s demand as well as projected needs of 1 Gbps service to each household or 10 Gbps for larger services to businesses to meet the communication requirements needed to enable municipal operations, connect public facilities or anchor institutions (libraries, healthcare facilities, public safety stations). The Commission may also want to consider new demand generated by the open-access middle-mile network such as new digital businesses that may be established as a result of the provision of service in currently underserved rural or tribal areas.

Establishing demand based on bandwidth and quality of service (including upstream and low latency) may address evolving uses and applications of broadband. The Commission should also consider computing simultaneous aggregate bandwidth demand during the busy hour to ensure the middle mile does not become the bottleneck for delivery of last-mile services.

Should the network also deploy additional conduit within each route for potential future expansion?

SANDAG is partnering with Caltrans and the County of San Diego to integrate broadband infrastructure along 18 miles of the SR 67 corridor. SANDAG is currently conducting a competitive procurement to select a broadband provider that can leverage this infrastructure to enhance broadband service available to communities along this corridor. As part of the project scope, the public agency partners are funding multiple conduits that can be used to pull fiber to serve Caltrans and County communication needs. This also includes the provision of extra empty conduits that can be leveraged by a public agency partner or ISP to pull fiber for broadband service to serve current and future needs. Consistent with our earlier response, SANDAG urges the Commission to employ a similar approach so that potential future expansions can be readily accommodated without a significant additional investment in public funding. Based on our

experience, SANDAG recommends that empty conduit banks be sized to account for the largest count fiber strand required to serve the communities and provide a minimum of four ducts of the same size, no smaller than the 1.5-inch inner diameter. Providing additional conduit within each route would enable the Commission, ISPs, or other agencies to seamlessly connect to the middle-mile network, future-proof the system, and create resiliency in the network.

Should these factors change based on the population density and distance from the core network?

While population density and distance are important factors for the CPUC to consider, they should not be sole determinants. As stated earlier, SANDAG recommends that the network be designed to provide 1 Gbps service to each household or 10 Gbps for larger services to businesses. This should also consider current and future demand driven by growth in population, jobs, or new demand induced by the open-access middle-mile network.

III.                  Conclusions

SANDAG thanks the Commission for consideration of these comments and looks forward to the partnership with the Commission and other stakeholders to support the expansion of the open-access middle-mile network. Building collaborative and transparent partnerships with public and private stakeholders is critical to the vision outlined in the State’s Broadband Action Plan.

SANDAG is committed to expanding broadband service to the region’s unserved and unserved areas. We encourage the Commission to ensure there continue to be adequate opportunities for coordination with regional governments like SANDAG, broadband consortia, and local jurisdictions to maximize our public investments and meet mutual objectives to bridge the digital divide.

Dated: September 3, 2021, Respectfully submitted,

  /s/ Kirk Blackburn Kirk Blackburn

Legislative & Regulatory Advocate Ellison Wilson Advocacy, LLC

1201 K Street, Suite 1201

Sacramento, CA 95814

Tel: (916) 448-2187

E-mail: lobby@ellisonwilson.com

Advocate for SANDAG

Input to the CPUC on SB156 and Middle-Mile Network Rulemaking September 2021

California Emerging Technology Fund

Input to the CPUC on SB156 and Middle-Mile Network Rulemaking September 2021

1. Identifying Existing Middle Mile Infrastructure:

Attachment A provides a list of the state routes proposed for the statewide open access middle mile network, referred to as the “Anchor Build Fiber Highways.” These routes may also be viewed on an ArcGIS map, which can be found here: https://www.arcgis.com/home/webmap/viewer.html?w ebmap=e17e4e1c88b04792ab0a2c50aa1a19a3&extent=- 126.1445,34.5234,-

113.5981,41.1113

• What routes, if any, should be modified, removed from consideration, or revised? Provide an explanation for these suggestions.

CETF

The CPUC process to identify priority State Routes for an open-access middle-mile network should start first by comparing the list of candidate routes in Attachment A with the Strategic Broadband Corridors Report prepared by the Regional Consortia and submitted to

the California Broadband Council in November 2018. The Strategic Broadband Corridors were identified through an open consultation process with Regional Transportation Agencies, coordinated by the California Association of Councils of Governments, under the umbrella of California Forward in cooperation with the California Department of Transportation (Caltrans). Importantly, the Broadband Strategic Corridors were identified and prioritized based on reaching unserved households, which should remain the primary criterion for State investment in government-owned middle-mile infrastructure. All middle-mile investments should be driven by a priority focus on reaching last-mile unserved households, especially high-poverty areas and Tribal Lands.

For several Regional Consortia, identification of Broadband Strategic Corridors was based upon their work to prepare Preferred Scenarios to achieve ubiquitous deployment at scale throughout their region, thereby assisting the CPUC in meeting the State’s statutory goal of achieving at least 98% in all regions by 2022 (although that goal is far from being achieved and proposed bills (SB4) to extend collections into the California Advanced Services Fund (CASF) move that deadline far into the future. Although the Regional Consortia used CPUC maps with the previous definition of unserved (10/1 Mbps), the Preferred Scenarios remain

viable because they focused on getting to the hardest-to-reach households, which means that all newly-defined unserved households at 25/3 Mbps and all anchor institutions are passed along the path of deployment. Further, the Preferred Scenarios planned to reach 100% of all unserved households, which is the strategic approach to be assured of achieving at least 98%.

Also in 2018 Caltrans and the California Transportation Commission (CTC) adopted updated guidelines for transportation corridor planning that recognize “broadband as a green

strategy” to improve mobility and reduce impacts on the environment. These transportation guidelines are practical tools in advancing the notion of “Dig One, Dig Smart” policies and practices because they encourage the incorporation of broadband into transportation projects for economies of scale, not just use transportation corridors rights-of-way (ROWs) to build government-owned middle-mile broadband networks. The CPUC should advocate and the Department of Technology (CDT) must ensure that the Third-Party Administrator (TPA) engaged to oversee construction of the middle-mile network actually incorporates the spirit and intent of “Dig Once Dig Smart” policies and practices. This means that another prioritization of the middle miles by the CPUC, CDT, and TPA must come from taking into consideration: (a) all planned transportation projects (including scheduled maintenance resurfacing and overlay projects; and (b) all Caltrans priority corridors for intelligent transportation systems (ITS) for traffic controls. There also are segments of the State’s transportation network for which conduit was installed at the time of construction to facilitate the deployment of broadband, including H99 in Merced County and SR198 in Kings and Tulare Counties. These segments with existing conduits are assets to consider as another factor in prioritizing deployment.

While the government-owned middle-mile network is envisioned to align primarily with the State’s transportation network, which is a strategy that CETF has advocated for more than a decade, there are other ROWs and alignments that should be considered, particularly the State Water Project, Irrigation and Water Districts, and energy utilities. For example, in Imperial County, the Imperial Irrigation District (IID) owns as many vital ROWs as Caltrans. Fortuitously, the Southern Border Broadband Consortium (managed by Imperial Valley Economic Development Corporation) secured from IID a willingness to consider collaboration in conjunction with the preparation of the Imperial County Preferred Scenario. Another example of substantial planning with the explicit engagement of an investor-owned utility (IOU) was led by Riverside County with the cooperation of all 28 cities.

Finally, there are pending applications before the CPUC that will provide the critical middle-mile infrastructure that should be approved, several of which should have been expedited and approved years ago, such as the Northeast Loop for 5 Counties along SR299, SR139, SR36, and the Kern Valley Project along SR178, SR14.  Deployment of broadband infrastructure along H299 from Eureka to Redding to Alturas is obvious and it is most regrettable that the Eureka to Redding project (D299) has not been built since the establishment of the CASF program (the Redding to Alturas segment is part of the Northeast Loop Project negotiated by CETF in the MOU with Frontier). Further, all of the pending projects for the Redwood Coast Region will establish vital middle-mile segments.

In addition to identifying State Routes for middle-mile deployment, there are some unique situations off the coast of California that should be constructed for middle-mile backhauls, such as an underwater fiber cable to Catalina Island and the City of Avalon. The City of Avalon and civic organizations already are working with ISPs to upgrade the network on the Island, but need a fiber connection to the mainland to promote economic vitality and ensure public safety.

It must be underscored that the State investment in a middle-mile government-owned network needs to be approached by CPUC, CDT and TPA with an intensity of focus and sustained, engaged collaboration akin to the Manhattan Project, but with full openness and transparency. It is essential that CDT, TPA, CPUC work with and through existing structures and ongoing efforts, especially the Regional Consortia and leading Metropolitan Planning Organizations (MPOs), such as the Southern California Association of Governments (SCAG) and San Diego Association of Governments (SANDAG) under the joint collaborative umbrella of Southern California Transformation. SANDAG and SCAG are providing trailblazing leadership to achieve Digital Equity and are working with their Regional Consortia. The State must respect and incorporate their recommendations.

There also are Local Governments that have taken the initiative to accelerate broadband deployment and adoption, such as the City of San Jose, the City of Los Angeles, City of Fresno, South Bay Cities Association in the SCAG Region, County of Los Angeles, County of Nevada, County of Tuolumne, and County of Ventura. Other Local Governments will be stepping forward as a result of the historic State investment in broadband, which is leadership that must be embraced and incorporated into the middle-mile planning. Outside of Southern California, CPUC, CDT, TPA should request and rely upon the Regional Consortia to convene all of the Local Governments in their regions to provide input on priorities for middle-mile deployment, which they did previously in identifying Broadband Strategic Corridors in 2018.

Once the above work has been completed to prioritize essential middle-mile infrastructure to reach all unserved households and Tribal Land, then CDT and the TPA should issue an open, competitive “Request for Partnerships” (RFP) to determine which existing Internet Service Providers (ISPs), both private and public, are willing to step up to provide access to existing middle-mile and/or build the missing middle-mile segments to reach last-mile unserved hardest-to-reach households. This approach ensures transparency and fairness in determining which ISPs are willing to help the State achieve ubiquitous broadband deployment while avoiding unnecessary duplication of middle-mile infrastructure. It should

be noted that SANDAG and SCAG are jointly developing an RFP that can serve as an example. The MPOs’ RFPs will include a Map of Needs and Opportunities with layers of data overlaid on the CPUC Broadband Map, including high-poverty areas, anchor institutions, and public assets, for ISPs to explicitly declare willingness and ability to step up. It is intended that the RFP will be structured legally such that those ISPs that do not respond to the RFP will have voluntarily and officially “stepped aside” without rights to future challenges.

In summary, the past work and existing efforts are foundational to jump-start planning the middle-mile network instead of reinventing wheels. All investments in constructing government-owned middle-mile infrastructure must be prioritized to reach unconnected households, with special attention to high-poverty areas and Tribal Lands. Further, the most cost-effective strategy is to focus on planning deployment to the hardest-to-reach unserved households, including all Tribal Lands, and then connect all other locations along the path of deployment. Any other approach will sub-optimize State investments and waste funds that otherwise could be used to reach last-mile unserved households.

• Are there existing middle mile routes that are open access, with sufficient capacity, and at affordable rates on the county highway routes listed in Attachment A?

CETF

All projects built with American Recovery and Reinvestment Act (ARRA) grants, such as D395 and CVIN, already should be open access and the State should not have to further subsidize these middle-mile projects. See attached position on open access.

In fact, these projects are good examples that show “if you build it, they will not come—at least not quickly”—there remains underutilization of these assets. There must be a focus on last-mile unserved households from the very beginning in collaboration with all public and private stakeholders to ensure middle-mile investments will accelerate last-mile deployment and connections for households and anchor institutions (see attached the addendum that CETF proposed to all ISPs for the May 4, 2020, CASF applications, which Frontier did include for the Northeast Loop Phase II).

CETF has long advocated that all middle-mile construction subsidized by CASF should be open access, which is a condition that should be applied to all pending applications. CETF does not support open access for CASF-subsidized last-mile infrastructure. However, if the pending CASF applications were immediately approved with open access requirements for the middle-mile segments, the State can have an immediate win for expanding middle-mile back-haul capacity throughout California.

• In the context of these comments, what is sufficient capacity and affordable rates?

CETF

The CPUC, CDT, and TPA should determine “sufficient capacity” for any given segment of middle-mile infrastructure based on the number of unserved last-mile households and anchor institutions along the path of deployment that will be supported by the middle-mile backhaul. Then, there should be a reasonable “margin of safety” capacity added to each segment for public safety, future growth, and redundancy purposes. It is in planning for this “margin of safety” that expeditious and cost-effective public-private partnerships can be negotiated with ISPs regarding existing middle-mile infrastructure to avoid unnecessary duplication of middle-mile infrastructure.

The concept of “affordable rates,” while a complex matter, should not be overcomplicated. Affordable rates for open access to middle-mile infrastructure should be based on 2 primary factors: (1) amortization of new capital investment over the life of the infrastructure; and

(2) cost of operation and maintenance of the infrastructure over the life of the new infrastructure. The CPUC, CDT, and TPA should calculate an average baseline “affordable

rate” (perhaps by region) and make this public knowledge as a “benchmark” for negotiating cost-effective public-private partnerships. It must be a simple, straightforward value proposition in the public interest. Of course, the State always can make the open access rates lower by reducing the amount of new capital investment to be amortized, even

reducing that figure to “zero” as a matter of public policy. However, the actual embedded full-cost calculation should be made public to fairly evaluate opportunities for public-private partnerships with ISPs.

• For routes that are identified as being open access, with sufficient capacity, and at affordable rates, how should the Commission verify these claims (e.g., should Communications Division send a data request for service term sheets, rates, approximate dark fiber, lit fiber, and conduit capacity, etc.)?

CETF

The “Request for Partnership” (RFP) process referenced above is an excellent mechanism to obtain the proprietary information, subject to non-disclosure agreements (NDAs), to verify all ISP claims regarding ability and willingness to reach last-mile unserved households, especially high-poverty areas and Tribal Lands. The RFP should be structured as a legally-

binding “Step Up or Step Aside” invitation, such that those ISPs that do not step up have by definition no challenge opportunities to any government-owned middle-mile deployment. CPUC, CDT, TPA then only have to verify the information submitted pursuant to the RFP and subject to NDA.

Please note, however, that it is essential that the work described in the CETF response to Question 1 must be completed as the background and foundation for an RFP. The equivalent of the SANDAG-SCAG Maps of Needs and Opportunities should be part of the CPUC-CDT-TPA Request for Partnerships.

Further, it is possible through the RFP process that CPUC-CDT-TPA can determine if an ISP with existing middle-mile infrastructure can and is willing to make excess capacity available to the State in comparison to the baseline-benchmarked costs discussed above for construction of new middle-mile infrastructure, even if the ISP declines in the RFP process to participate in deployment to last-mile unserved households. In this case, the ISP representations can become part of an agreement with bonding for performance standards and hold harmless provisions compelling upgrades if necessary,

• Are there any other criteria that should be used to verify these claims?

CETF

Responses to the RFP should request documentation and independent verification of claims by credible parties identified by CPUC-CDT-TPA. In addition, assertions about capacity and ability should be backed up by performance bonds and hold harmless provisions in awarding funds and negotiating agreements.

• Priority Areas

Federal funding must be encumbered and spent in a limited time period. Additionally, unserved and underserved areas of the state are in substantial need of broadband infrastructure investment.

• Is it reasonable to assume counties with a disproportionately high number of unserved households (e.g., 50% or more unserved at 100 Mbps download) are areas with insufficient middle-mile network access?

CETF

No, this is not a reasonable assumption. As explained earlier, all middle-mile investments need to be driven by hardest-to-reach last-mile unserved households, especially high-poverty areas and Tribal Lands. The new definition of unserved simply masks the need to drive to all the areas that historically have been unserved. The new definition of unserved identifies areas to be upgraded along the path of deployment. But, unless the CPUC-CDT- TPA focus on the hardest-to-reach unserved households, the middle-mile investments risk becoming “middle miles to nowhere” with slower incremental progress toward reaching the most digitally disadvantaged rather than deploying infrastructure at scale which will accelerate access to high-speed Internet infrastructure for everyone, especially the most digitally-disadvantaged residents today.

• What other indicators, if any, should the Commission use to identify priority statewide open-access middle-mile broadband network locations (i.e., built expeditiously, areas with no known middle-mile network access, regions underserved by middle-mile networks, regions without sufficient capacity to meet future middle-mile needs)?

CETF

As previously stated and explained, the overriding criteria for prioritizing State investments in government-owned middle-mile networks is to support last-mile deployment to the hardest-to-reach unserved households. This approach (with the RFP process) is the most cost-effective strategy to accelerate deployment and provide adequate bandwidth for all locations along the path of deployment. At the end of the day, the metric that counts more than anything else is: How many of the hardest-to-reach households that previously had no access to high-speed Internet service have been connected, and how many households were in high-poverty areas and Tribal Lands. Everything else being referenced as indicators of progress are useful milestones, but they are only “inputs” to “outcomes”. Real accountability for results (measurable “outcomes”) needs to be built into the middle-mile initiative from the very beginning. If this is done right, then those measurable outcomes become the indicators for prioritization and the program is much more transparent.

However, if for some reason there is a middle-mile segment needed to enhance public safety in a location that isn’t being passed by the above approach (which is unlikely but theoretically possible), then government-owned middle-mile infrastructure investment for public safety purposes is entirely appropriate and a companion criterion to reaching last-mile unserved households.

• Assessing the Affordability of Middle Mile Infrastructure

A key consideration is determining the cost of various middle-mile services. Through identifying the costs of these services in California, as well as across the country and globe the Commission can identify a threshold whereby services can be considered reasonably affordable.

• What are existing providers paying or charging for middle mile services?

CETF

In addition to CPUC, CDT, and TPA doing research for comparisons to the baseline- benchmarked costs in California, this information can be obtained through the RFP process described herein.

However, there is a factor that CETF has flagged repeatedly that is likely to drive up the costs of new construction—for both middle-mile and last-mile deployment—which is the lack of skilled workers in California to meet the demand. Thus, the State needs to join forces with labor leaders, employer organizations, and community-based organizations (CBOs) to recruit and train digitally disadvantaged residents to help build the infrastructure that will benefit their own families and communities.

Although it should be obvious, it is worth making a differentiation between the cost for open access of middle-mile backhaul infrastructure as discussed herein and affordable broadband rates for last-mile low-income households. If the costs of new middle-mile construction are calculated as recommended, then the cost for last-mile Internet service for low-income households should not be impacted and the existing affordable offers, including the federal Emergency Broadband Benefit (EBB) Program (and its proposed successor Affordable Connectivity Program), should be sufficient. Further, if the CPUC and FCC Lifeline programs can ever be structured to work effectively, then affordability for last-mile low-income households is addressed through means other than the cost of new middle-mile infrastructure.

• Are there other factors or sources of information the Commission should consider for determining whether these services are affordable?

CETF

The CPUC should request information from other states and through the Federal Communications Commission (FCC), Department of Commerce National Telecommunications and Information Administration (NTIA), National Association of Regulatory Utility Commissions (NARUC), and ISPs.

• Is it reasonable for the costs of these services to change depending on the location where the service is provided (i.e., rural vs urban)?

CETF

Construction costs vary widely in California, which is why the baseline-benchmark process recommended above suggests doing that analysis by region.

• Leasing Existing Infrastructure

Indefeasible Rights of Use (IRUs) are long-term leases (generally 20 to 30 years) for unrestricted, legal capacity on a communications network for a specified period of time.

These contracts generally obligate the purchaser to pay a portion of the operating costs, and the costs of maintaining the infrastructure.

• If there is existing open access communications infrastructure with sufficient capacity to meet the state’s needs, should the state purchase IRUs from that network?

CETF

The RFP process is an effective approach to obtaining this information in addition to gathering information from FCC, NTIA, and ISPs.

• Is there any value in the state purchasing an IRU from the network if capacity is already available?

CETF

It should be a straightforward determination of what’s in the public interest as to whether or not there is value in purchasing an IRU from a network if capacity is already available.

CETF described that approach above. However, the State should not have to be paying for existing ARRA middle-mile projects that already have an open-access requirement.

• If the state relies on IRUs for the development of the statewide network, will the generational investment that this funding provides be diminished when the IRU leases end 20 to 30 years later? Will existing networks run out of spare capacity?

CETF

The answer to this question can be determined by calculating the baseline-benchmarked costs and analyzing the responses to the RFP.

• Interconnection: The statewide network will need to connect with other networks in order to deliver services.

• At what points should the statewide network interconnect (e.g., to other networks, servers, etc.)?

• Are additional exchange points necessary or strategic, and if so, where?

CETF

Issues regarding the appropriate network interconnections and exchange points for a government-owned middle-mile network (and for any IRU or pending CASF project) should be determined by starting with the focus to drive to the hardest-to-reach last-mile unserved households, especially high-poverty areas and Tribal Lands. Interconnections and exchange points are all a function of the needs and volume of demand by last-mile users.

• Network Route Capacity: The state will need to determine the amount of capacity to build into the network to meet existing and future demand.

• How many strands of fiber should the network deploy for each route?

• Are there other requirements or standards the Commission needs to consider to determine sufficient capacity?

• Should the network also deploy additional conduit within each route for potential future expansion?

• Should these factors change based on the population density and distance from the core network?

CETF

The answers to these questions will be derived from (a) prioritization of middle-mile investments to drive to the hardest-to-reach last-mile unserved households, especially high-poverty areas and Tribal Lands coupled with (b) Request for Partnership process to invite

ISPs to “Step Up or Step Aside” as delineated above. This approach accelerates deployment of high-speed Internet for all residents, instead of incremental builds to pockets of newly-defined unserved areas, and it obtains all the data and proprietary information needed to make the most cost-effective investments in middle-mile infrastructure in the public interest.

Attachments

1. Regional Consortia Report on Broadband Strategic Corridors
2. BroadbandUSA Fact Sheet on Open Access for ARRA Projects
3. Proposed Addendum Regarding Anchor Institutions for CASF Applications