Impact of CAF II–Funded Networks

America is plagued by a stubborn and growing rural-urban digital divide: Thirty-nine percent of rural areas lack access to broadband of 25 Mbps download/ 3 Mbps upload, compared with 4 percent of urban areas. This opportunity gap is a significant barrier for people who live, work and learn in rural communities. In contrast, communities served by gigabit networks experience clear economic benefits.

Making investments to close these ruralurban gaps can be a challenge for large telephone companies. The returns on investment in rural exchanges are rarely competitive with those in more densely populated areas, as high costs for network infrastructure often outweigh possible short-term economic returns from a small, dispersed customer base.

The FCC’s Connect America Fund (CAF II) program is designed to close the financial gap so large carriers, regulated as price-cap carriers, can improve broadband services in areas otherwise too expensive to serve. These carriers have accepted payments from the FCC to deploy broadband services to a targeted number of households in their service areas. Over the five-year program, four companies serving rural Minnesota will receive federal funds for broadband deployment. Construction using CAF II dollars began in 2016 and must be complete by the end of 2020. Service providers must offer broadband at speeds of at least 10 Mbps downstream and 1 Mbps upstream and meet other performance and pricing requirements.

But is CAF II funding adequate to encourage providers to deploy networks in less densely populated exchanges that deliver services comparable to the networks they build in more densely populated areas?

There is reason to believe the funding is not adequate. Though new FTTP networks provide the best infrastructure platform for robust rural connectivity, rural fiber deployments can range from $4,000 to more than $10,000 per passing, depending on population density and terrain. However, in Minnesota, CenturyLink is receiving $2,354 per household and Frontier $2,936 per household.

A STUDY OF TWO EXCHANGES

This article examines CAF II–funded broadband network deployments in two rural Minnesota telephone exchanges, one (Lindstrom) owned by Frontier and the other (Braham) by CenturyLink. These exchanges contain characteristics seen throughout the state. Both are quite rural, with small cities, farms, exurban and rural homes, wetlands and lakes. Of the two, the Braham exchange is the more rural and less densely populated. 

Community Technology Advisors conducted the primary field research for this study in the summer of 2017. Using base maps and a GPS-enabled camera to take pictures of various electronics boxes located in public rights-of-way, we mapped CAF II–funded improvements within the two exchanges. Electronics boxes clearly identified as electric utility, railroads or traffic/lighting control devices were not mapped.

Where boxes could not be definitively identified, we assumed they were digital subscriber line access multiplexers (DSLAMs), possibly overstating the deployment of electronics and the resulting service areas. DSLAMS are the field electronics that connect newly deployed fiber lines and the existing copper lines that connect to customers.

Because neither provider responded to requests for comments and corrections on this report, our conclusions are based on observations of field deployments.

Prospective service areas surrounding the DSLAMs are indicated with 3,000- foot radius and 9,000-foot radius circles. It is likely that customers within 3,000 feet of a DSLAM will be able to receive broadband services at or above 25 Mbps/3 Mbps; those within 9,000 feet of the DSLAM will be able to receive at least 10 Mbps/1 Mbps.

Studies of upgrades to two Minnesota exchanges show that carriers are installing fiber to the node, and few households will have access to broadband speeds of 25 Mbps/3 Mbps.

Possible limitations in this approach include the following:

1. Copper lines do not radiate directly from a DSLAM but follow the roads. Each turn adds distance that limits deliverable capacity. 
2. Broadband speeds vary greatly depending on the condition of the copper lines. It would not be unusual for copper lines in rural areas to be at or even beyond the ends of their expected useful lives. Conversely, copper lines could be of higher quality than assumed.
3. One or more DSLAMs could have been missed in this field inventory.
4. Spare copper pairs that would allow two or more pairs of copper lines to be bonded together may be available.
5. Ongoing technical improvements in electronics and software could allow for future improvements in DSL capabilities.

Despite these uncertainties, we believe this methodology provides a reasonably accurate picture of the impact of these deployments. This confidence is backed up by ongoing conversations with rural broadband customers across the state in community and regional meetings, with many interactions documented through emails and video testimonials.

UNDERSTANDING THE NETWORKS

The CAF II–funded, fiber-connected electronics within these service areas are “nodes” or DSLAMs connected by fiber to the providers’ central offices; existing copper telephone lines serve as the last mile (or more) from nodes to customer premises. The quality of internet service over copper degrades with distance. That means homes closer to the node will have faster service.

Homes or businesses within 3,000 feet of a node can generally experience sustained speeds of 25 Mbps/3 Mbps or more. Properties within 9,000 feet can generally experience speeds of at least 10 Mbps/1 Mbps. Beyond 9,000 feet, properties generally will experience speeds slower than 10 Mbps/1 Mbps.

The maps of the two exchanges suggest two challenges for communities in CAF II areas:

• First, the newly built CAF II– funded networks do not provide service that meets Minnesota’s state speed goals, which are 25 Mbps/3 Mbps by 2022 and 100 Mbps/20 Mbps by 2026, to all households.
• Second, the lack of transparency from CAF II recipients about their network deployment plans makes it difficult for communities to do their own planning and leverage public investment.

FRONTIER’S LINDSTROM EXCHANGE

The Lindstrom exchange includes most of Chisago Lakes, a scenic area straddling U.S. Highway 8, including the towns of Chisago City, Lindstrom, Center City and Shafer. The community has a small-town or exurban feel; it has an excellent school system and easy access to Twin Cities metropolitan area work opportunities. The numerous lakes, wetlands, woods and small farms make living outside the boundaries of incorporated cities attractive. The rural residents are widely and generally evenly dispersed across the exchange, with higher densities around the many recreational lakes.

In a 2015 survey of more than 800 county households,

• 27 percent said they used the internet to operate their businesses.
• 31 percent said they would start home businesses with better internet.
• 35 percent said they would telecommute with better internet service.
• 45 percent said they would use the internet for educational purposes.
• 86 percent said they would use the internet for all kinds of purposes if it was available.

Each Frontier DSLAM connected by fiber optic cabling was verified based on the easily identifiable fiber optic cable pole markers. Frontier fiber was generally installed along each paved county road in the exchange. For the most part, Frontier utilized a standardized equipment installation throughout the exchange, with a DSLAM on the right, cross-connect in the center and power meter on the left.

To see the impact of CAF II investments toward meeting the 2022 state broadband goal of 25 Mbps/3 Mbps, view the 3,000-foot-radius circles on the map. As noted earlier, customers within 3,000 feet of a fiber node can receive services at the 25 Mbps/3 Mbps level and possibly much faster if a provider deploys either vectoring or pair bonding technology.

There is another prospective opportunity for vastly improved services – direct fiber connections – for customers who seek higher levels of broadband services. This option, in the Lindstrom exchange or in future deployments by any provider, depends on the proximity of a customer to the fiber routes and the provider’s network design. If handholes (which provide easy access to fiber underground in the same way manholes provide access to tunnels under roads) were installed along the route between the nodes, especially at driveways leading to farms, rural businesses or high-end homes, a provider could offer solutions to customers with high demands for bandwidth and a willingness to pay for installation and improved services. With fiber construction costs of $20,000 per mile, proximity to a handhole is critical.

At this time, it is unknown whether handholes were part of the network design. Industry experts consulted for this study note that such subsequent enhancements would add operational complexity to the network, ranging from electronics to technician training.

The map suggests that

• Approximately 10–15 percent of the land area is within a 3,000-foot circle and thus capable of receiving 25 Mbps/3 Mbps or better.
• Relatively few areas are outside the 9,000-foot-radius circles, where available service is likely to be less than 10 Mbps/1 Mbps.
• Most of the land within the exchange is within a 9,000-foot circle but outside a 3,000-foot circle, indicating that for most customers, broadband service availability would be between 10 Mbps/1 Mbps and 25 Mbps/3 Mbps.

Note that the study measured land area, not the number of potential customers. If customers are concentrated near the nodes, a higher percentage of customers could receive 25 Mbps/ 3 Mbps service.

The results of field examination of Frontier’s CAF II network deployment for the Lindstrom exchange (outlined on the map in gray). Each dot represents an identified DSLAM. Each red circle has a 3,000-foot radius; each blue circle has a 9,000-foot radius.

 

CENTURYLINK’S BRAHAM EXCHANGE

Braham is located in east-central Minnesota directly north of Minneapolis-St. Paul. The exchange straddles four counties – Chisago, Isanti, Kanabec and Pine. It is farther north than the Lindstrom exchange, considerably more rural (less densely populated) and observably less prosperous.

Though data is not available by exchange, city data indicates that both income and house values are lower here than in the Lindstrom area. Population growth is slower. Braham is classified as 100 percent rural, compared with Lindstrom, which is only 6 percent rural; there is little, if any, pressure here for suburban-like development. Because the distance to the Twin Cities is greater, residents are more likely to be employed locally in small regional center communities.

The exchange includes small towns (Braham is the largest community and has just under 2,000 people), wetlands, small lakes and relatively small farms. Significant numbers of people live in the rural countryside along county and township roads.

Though business development opportunities for CenturyLink in the Braham exchange may seem limited, the city of Braham recently partnered with a small wireless ISP to bring fiber-based internet services to its new business park in response to the requirements of a prospective tenant.

A broadband survey in Isanti County in fall/winter of 2017 suggests a need for network improvements.

CENTURYLINK’S INFRASTRUCTURE INVESTMENT STRATEGY

CenturyLink focused its CAF II infrastructure investments in the more densely populated portions of the Braham exchange. Most of the identified deployment is in the eastern and southern portions of the exchange.

Though new fiber routes were clearly marked with the traditional white and orange posts, large areas of the exchange seem to depend on buried copper lines. Based on the “buried cable” warning signs and the older pedestals, industry experts consulted for this study advised that fiber connectivity was highly unlikely.

A variety of equipment has been installed, seemingly dependent on population density. In the denser areas, DSLAMs were larger, with varying numbers of ports. In the less dense areas, 24-port DSLAMS were common.

In the field, we saw equipment that was new and some DSLAMs that appeared to be older and to have been installed in earlier years. In some areas west of Highway 65, DSLAMs appeared to be connected to copper lines rather than fiber. Given their apparent age, it is likely that these older DSLAMs might be fed by bonded T-1 lines with far less bandwidth capacity than new, fiber-fed DSLAMs.

It is also possible that some of these older DSLAMs were taken out of service, but not removed, when the new equipment was installed. This is most likely to have occurred in the areas where DSLAMs are quite close together.

The results of field examination of CenturyLink’s CAF II network deployment for the Braham exchange (outlined on the map in gray). Each dot represents a DSLAM (some may be inactive). Each red circle has a 3,000-foot radius; each blue circle has a 9,000-foot radius.

 

Compared with Frontier’s Lindstrom exchange, CenturyLink has numerous interlocking red circles, with DSLAMs located less than 1 mile apart. Unfortunately, it was not possible to determine which, if any, DSLAMs are now inactive.

• Significant portions of the Braham exchange are more than 9,000 feet from a node, indicating that internet service below 10 Mbps/1 Mbps would be available to residents and businesses in these locations.
• In the areas where CAF II investments were made, there is relatively consistent coverage within the 9,000-foot-radius circles. This suggests that, in these areas, customers should have access to sustained 10 Mbps/1 Mbps speeds.

The areas to receive broadband coverage of at least 25 Mbps/3 Mbps are relatively limited, based on CenturyLink’s CAF II–funded upgrades to date. Actual speeds in these areas could range up to 80 Mbps depending on proximity to electronics, vectoring and pair bonding. Vectoring is generally not effective at distances greater than 3,000 feet.

Without feedback from CenturyLink, it is unknown whether or not the carrier has completed its CAF II–funded deployments in this area. It is possible that CenturyLink will install additional fiber and electronics in the future.

However, the upgrades deployed to date leave many areas unserved. People living in areas left behind see their neighbors getting service but don’t know whether they too eventually will be served.

ARE 10/1 SPEEDS ADEQUATE?

This study raises two challenges for CAF II–eligible rural communities and policymakers: first, the adequacy of CAF II–funded networks to meet communities’ present and future needs; second, the need for greater transparency as to how, where and when the networks are built.

CAF II program critics have roundly derided the FCC’s 10 Mbps/1 Mbps requirement as inadequate at a time when the FCC’s own minimum broadband standard is 25 Mbps/3 Mbps and when providers in competitive markets are regularly delivering gigabit symmetrical service.

Doug Dawson, a nationally recognized telecommunications consultant and president of CCG Consulting, elaborates:
These upgrades will improve broadband in the affected areas, but only by a small amount. Some residents in these areas today can get very slow DSL, under 1 Mbps. There are also numerous WISPs (wireless internet service providers) operating in the area offering speeds under 5 Mbps. And everybody always has the option of satellite broadband, which is universally disliked due to the latency and data caps.
The really bad news for these areas is that this upgrade is going to be in place for a long time. The FCC is probably not going to think about the CAF II areas again until well past the end of the CAF timeline, perhaps not until 2025. By 2025, the average household in the country is going to probably want a 100 Mbps connection if the current broadband growth trends continue.

In response, the FCC pointed to the fact that the federal broadband standard was 4 Mbps/1 Mbps at the time the CAF II program was being designed. Defenders of the program also point out that the limited federal funds available should be used to provide some level of broadband to the largest number of households.

A few projects have combined state and CAF II funding to build networks that meet state goals for broadband service.

To address concerns about the capability of the technologies used in these CAF II–funded upgrades, internet service providers often point to new and emerging technologies that can greatly enhance the capacity of DSL, such as G.fast, vectoring and pair bonding. In the case of G.fast, however, a review of the online literature indicates that its successful deployment requires so much fiber that it is most suitable in urban settings with multiple-dwelling-unit buildings where fiber optic cabling is just outside the building.

Even if G.fast and vectoring worked well in the rural settings where CAF II networks are being built, upgrading the networks to incorporate these more advanced technologies would require additional investment by providers.

Another defense of the 10 Mbps/1 Mbps expansion is that it is good enough or at least better than nothing. Although the CAF II–funded upgrades do result in access to broadband services of at least 25 Mbps/3 Mbp for some rural residents, many more residents these upgrades serve will end up with access to speeds much slower than that.

The 10 Mbps/1 Mbps access requirements also does not satisfy Minnesota broadband goals. Accepting lower speeds for some parts of Minnesota creates second-class status for rural residents.

LACK OF TRANSPARENCY

Another issue with the CAF II program is its lack of transparency. CAF II does not require participating providers to share or report on any specifics in their network plans beyond the total number of households to be served.

The program includes some interim milestones, beginning with 40 percent of the recipients’ statewide build commitment being completed by the end of 2017. (CenturyLink reports it is on track to complete 60 percent of its CAF commitments by the end of 2018.) Policymakers have asked the FCC to improve transparency of CAF II plans and projects, so far to no effect.

This puts community leaders of unserved areas in a quandary. CAF II upgrades do not meet the state speed goals, but they are better than what residents currently can get. But how much better?

In public meetings and statements, CAF II recipients underscore their commitments to offer broadband speeds that meet or exceed the CAF II 10/1 requirements, often emphasizing the faster speeds available only closest to the nodes.

Nationally, CenturyLink claims that about 70 percent of the homes in its target areas served to date with CAF II–enabled networks have access to speeds of 20 Mbps or higher.

In Minnesota, community leaders and policymakers are evaluating these statements in the absence of an onthe-ground picture of what’s actually being built. This lack of clarity about how participating carriers are spending their CAF II funds inhibits planning and informed stewardship of public resources.

CONCLUSION

With the firm link established between quality broadband availability and community economic vitality, community leaders want to ensure that their residents, businesses and institutions are not left behind, whether downtown in the county seat or miles from town on a farm or a lake. Nextgeneration people and technologies will demand an even higher standard of broadband quality.

CAF II funding increases internet speeds to some customers, but for many the increase is not enough to meet the state’s broadband goals for either 2022 or 2026. In sum, CAF II investments in Minnesota are being spent to build networks that don’t meet today’s federal definition of broadband and won’t meet state goals for the future.

Moreover, lack of transparency in proposed CAF II network plans and timelines is making it difficult for impacted communities to plan accordingly to ensure their broadband needs are being adequately met.

Despite these challenges, Minnesota can boast examples of communities and CAF II recipients working together to finance networks that offer better service than CAF II–funded networks alone. Projects in Martin County and in Chisago County’s Sunrise and Fish Lake Townships have combined state and CAF II funding to build networks that meet state goals. To qualify for state grants, providers were required to commit to high-speed networks (25 Mbps/3 Mbps with scalability to 100 Mbps), far higher levels of service than the CAF II–funded projects in the Lindstrom and Braham exchanges.

Examples include the following:

• Frontier has committed to delivering ubiquitous 25 Mbps/3 Mbps service with densely deployed fiber and DSLAMs in Nobles County.
• CenturyLink, using bond funds from the township residents in addition to CAF II and DEED funds, is deploying a fiber-to-thehome network capable of gigabit speeds in Fish Lake and Sunrise Townships in Chisago County.

These projects demonstrate the potential of Minnesota’s nationally renowned Border to Border Broadband grant program as an effective tool for leveraging state funds to impose state requirements – geared toward state goals – to ensure maximum public benefit from public investments.

In addition, greater transparency on the part of CAF II recipients would benefit communities and providers alike. Though recognizing that plans sometimes change, maps showing planned CAF II deployments over the next three years (through the end of the program) would help communities partner with their providers for better networks, as in the examples cited previously. The same holds true of the FCC’s Alternative Connect America Cost Model (A-CAM) subsidies to medium-size rate-of-return carriers.

With greater transparency from CAF II recipients, state elected officials and the Office of Broadband Development could build effective strategies for maximizing the value of CAF II and A-CAM dollars for the benefit of Minnesotans, including designing programs that require a fair mix of public resources – federal, state and local – to spur even more and better rural deployment in a predictable way.

Maximizing the public benefit from public investments is good for everyone.

Better broadband through transparency and collaboration is possible.