Evolving Human Systems

This is the first in a series of blog posts outlining an approach to expanding high-speed Internet connectivity that I refer to as FAST PIIPs (Fiber and Spectrum Together for Public Interest IP Services).  The FAST PIIPS model reflects the evolution of my thinking about how to bridge remaining gaps in broadband access and adoption, including a 2016 collaboration with Michigan State University’s Quello Center on a report entitled Wireless Innovation for Last Mile Access: An Analysis of Cases and Business Strategies and a policy brief entitled Bridging Michigan’s Social and Digital Divides.

While the FAST PIIPS model can be applied to both rural and urban connectivity challenges, my focus in these posts will be on rural areas, where the challenges tend to be more widespread and severe, and where the new FCC chairman and his critics seem to share a sense of urgency, if not a strategy for how best to address it.  My hope is that this overview of the FAST PIIPS model will contribute in some small way to development of public policies and network investment strategies that do a better job of digitally connecting and empowering the nation’s economically and socially stressed rural communities.

Though I’ll get into more detail in later posts, I want to start by briefly describing the two main components of the FAST PIIPS model. The first element, “FAST” (Fiber and Spectrum Together) relates to technology and can be read to mean “extend fiber as far as you can afford, then use it to provide high-capacity backhaul to support wireless connections to locations that can’t currently be reached cost-effectively by fiber.”  While the determination of how far fiber can “affordably” be extended may vary considerably and be a subject of intense debate in specific situations, the basic network design principle is pretty straightforward and, I think, uncontroversial.

The second element of the model, “PIIPS” (Public Interest IP Services) may be a bit more controversial in that it is focused largely on the institutional arrangements and motivations related to deploying access and backhaul networks.  As I wrote on pg. 4 of the Wireless Innovation for Last Mile Access report and will discuss more in later posts:

“Given the economic and other challenges associated with bridging our nation’s remaining gaps in broadband access and the benefits it provides, locally-anchored enterprises and business models focused on sustainably addressing local needs may be better suited to this task than enterprises and business models focused on maximizing investor returns in national or global markets.”

That report, commissioned by Merit Network and five other state Research and Education Networks (RENs), examined a dozen examples of wireless innovation in last mile access.  As the paragraph cited above went on to note, “virtually all of the cases examined in th[e] report embod[ied] some institutional and business case elements reflecting this prioritization of community benefits versus shareholder returns,” with the range of institutions involved in these innovative projects including public libraries, universities, K-12 school districts, local governments, cooperatives, community nonprofits and privately owned “social enterprises.”

To close out this introductory post I’ll note that the FAST PIIPS model includes technology-related (i.e., FAST) elements that are especially supportive of the PIIPS component’s focus on directly targeting the public interest goal of expanding access versus pursuing it through private market dynamics. Chief among these are the use of: 1) low-cost unlicensed and education-focused spectrum not allocated via competitive auctions and; 2) open access fiber backhaul networks operated by mainly nonprofit entities and deployed with financial support from the federal government, as occurred under the Broadband Technology Opportunities Program (BTOP) funded by the American Revitalization and Reinvestment Act of 2009, a topic I discussed in this series of posts on the Quello Center blog.

Below are links to the other posts in this series.  Feedback from readers, especially constructive criticism, is welcome.

• Part 2: TV White Space as a Tool to Expand Rural Access
• Part 3: EBS Spectrum as a Tool to Bridge the Rural Homework Gap
• Part 4: Fiber and Spectrum Together (FAST) for Rural Access
• Part 5: The Role of Community Benefit-Focused Organizations
• Part 6: Research to Support Success

In an earlier post I discussed the potential of TV White Space (TVWS) spectrum to help expand the availability of affordable high-speed Internet access in rural areas.  In this post I want to bring another underutilized spectrum band into the rural access equation: Educational Broadband Service, or EBS.

While there’s a lot that can be said about EBS spectrum and its history (for example, see Sec. 2.1 of this paper by Aleksandr Yankelevich) this post will focus on several factors that make EBS a potentially useful tool for addressing the rural broadband challenge as part of a FAST PIIPS (Fiber and Spectrum Together for Public Interest IP Services) connectivity strategy:

1.  EBS spectrum includes five 22.5 MHz spectrum blocks totaling 112.5 MHz in the 2.5 GHz band. Operating in this band enables EBS networks to leverage technical progress and volume-related cost reduction already achieved in the large and well established global LTE equipment market. Though EBS lacks the strong non-line-of-sight capabilities of TVWS, EBS networks can operate at higher power levels than networks relying on unlicensed TVWS or Wi-Fi technologies.  And, as users of exclusively licensed spectrum, they are not subject to the type of interference constraints that, as demand intensifies, can disrupt performance and business models in unforeseen and uncontrollable ways for networks relying on unlicensed spectrum.
2. EBS spectrum is licensed only to educational institutions, which positions it especially well to serve educational purposes, including helping to bridge the “homework gap” that has become an increasingly important issue within the educational community, especially with regard to poorly connected rural areas.
3. Much of today’s licensed EBS spectrum, especially in metro markets, is currently leased on a long-term basis to Sprint and other commercial wireless service providers.  As a result, its education-focused value in these markets is limited mainly to using lease payments to support educational purposes and programs.
4. In contrast to these metro markets, licenses for EBS spectrum in some rural areas have not been licensed by the FCC, and thus lays dormant.  In recent years this has led a handful of pioneering educational institutions to request and receive a waiver to use unlicensed EBS spectrum to deploy LTE-based education-focused networks.

Michigan pioneers use of EBS to expand rural connectivity

To my knowledge, the most ambitious example of this waiver-based use of EBS spectrum is Northern Michigan University (NMU), which the FCC has authorized to use the full 112.5 MHz of EBS spectrum to serve the state’s thinly populated Upper Peninsula. Last November NMU began offering its Educational Access Network (EAN) service through other educational institutions in the region and directly to individuals seeking improved access to online educational resources.  In late March of this year the university announced receipt of a $6.5 million state grant from the Michigan Strategic Fund Board, which it will match with $3.2 million of its own funds to expand EAN’s existing coverage area to 64 cities and townships in the Upper Peninsula over a two-year period.

According to its web site, the EAN offers “best effort” off-campus connectivity using carrier-grade wireless equipment at a range of speeds and prices, none of which is constrained by monthly data usage limits.  Tiers include $19.95/mo. for 5Mbps/2Mbps service, $24.95 for 10Mbps/5Mbps, and $34.95 for a 15Mbps/5Mbps tier. The EAN site also notes that some “degree-seeking undergraduate and graduate students receive access to the EAN as part of their tuition and fees.”

Depending on distance from a transmitter and obstacles impeding line-of-sight, EAN end-user equipment options include portable LTE devices (usable where signal strength is strongest), indoor units, or outdoor antennas. The latter provide the strongest connections and are presumably necessary to obtain reliable service at the edges of network coverage. Based on the address a user provides upon enrollment, NMU provides recommendations regarding the type of connectivity devices best suited for that location.

Though the EAN network is currently licensed to serve only Michigan’s Upper Peninsula, the above map shows that other rural portions of the state also have a substantial amount of unlicensed EBS spectrum that could be used to support an expansion of EAN or launch of another education-focused network (this county-based map provides only an approximation, since EBS licenses are currently issued based on a 35-mile coverage radius).

As the map indicates, 3-5 EBS spectrum blocks (67.5-112.5 MHz) remain unlicensed–and therefore potentially available via waiver or renewed FCC licensing–in much of the northern (and generally rural) portion of Michigan, whereas most of the southern portion of the state, where its larger and more densely populated cities are located, has no EBS spectrum available.

In my next post I’ll consider the overlaps between available TVWS and EBS spectrum, using a handful of small Michigan communities as examples. I’ll also consider how this combined spectrum availability can be combined with the increased availability in rural areas of open access high-capacity fiber backhaul, much of it funded through federal programs such as the stimulus-financed Broadband Technology Opportunities Program and the E-Rate program that has helped finance fiber connectivity for many schools and libraries. This extension of mainly nonprofit fiber to CAIs and other locations in smaller communities and rural areas can provide last mile networks–whether wireless or wired–with increased bandwidth and improved economics when it comes to the backhaul capacity needed to support reliable and scalable high-speed connectivity within their communities.

Below are links to the other posts in this series. Feedback from readers, especially constructive criticism, is welcome.

• Part 1: Introduction
• Part 2: TV White Space as a Tool to Expand Rural Access
• Part 4: Fiber and Spectrum Together (FAST) for Rural Access
• Part 5: The Role of Community Benefit-Focused Organizations
• Part 6: Research to Support Success

In my last two posts I discussed how unlicensed TVWS spectrum and education-focused EBS spectrum are underutilized and potentially valuable connectivity resources in many underconnected rural communities. In this post and the next, I hope to flesh out my vision of a rural FAST PIIPS model a bit more, with a focus on the connectivity potential offered by these two spectrum bands when combined with open-access non-profit high-capacity fiber backhaul.

As I did in those two initial posts, I’m going to use Michigan as an example, mainly because it’s the state I’m most familiar with, as both a resident and policy analyst.  As a first step, I’m going to compare the amount of unlicensed TVWS and EBS spectrum in seven small communities in northern Michigan to the average sub-3 GHz spectrum holdings of the nation’s four dominant mobile operators following this year’s 600 MHz auction, according to data compiled by research firm Allnet Insights & Analytics.

As the above graph clearly shows, the combined total of  unused TVWS (the blue <1 GHz bar) and EBS (the green 2-3 GHz bar) spectrum in all seven of these small Michigan communities is far greater than the nationwide average holdings in comparable spectrum bands for each of the four major mobile carriers.  While the latter ranges from 110 to 202 MHz and averages 146 MHz, the potential sub-3 GHz FAST PIIPS spectrum in the seven rural communities ranges from 230 to 329 MHz and averages 289 MHz, nearly twice the average for the four dominant national wireless carriers.

While these spectrum counts do not provide clear apples-to-apples comparisons, they do suggest that TVWS and EBS spectrum have potential to deliver digitally-empowering data rates in sparsely populated rural areas.  I was recently involved in a research project examining pioneering projects exploring this potential, and would strongly encourage further research in this area, as well as additional (and well-documented) deployments that learn from these pioneering projects and leverage the continued evolution of technology and business models related to addressing the rural connectivity challenge.

Fiber backhaul: The “F” in FAST PIIPS that helps ensure it will be “fast”

As noted briefly above and in the two preceding posts, the use of abundant and low-cost yet underutilized spectrum to support last mile access is not the only technology component of the FAST PIIPS connectivity model.  Another core element is the availability of open access, high capacity fiber backhaul.

In many states, an increasing amount of this fiber backhaul capacity has become available this decade as mainly-nonprofit network operators have extended fiber deeper into rural areas with support from the Broadband Technology Opportunities Program (BTOP) established by the American Recovery and Reinvestment Act of 2009.

In Michigan, the primary recipient of BTOP infrastructure funding was Merit Network, a nonprofit entity that mainly serves the state’s educational community and other community anchor institutions (CAIs), and is owned by Michigan’s major universities.

The red and blue lines in the above map (original here) show how Merit’s network expanded more deeply into Michigan’s rural areas during two waves of BTOP-supported network construction in the first half of this decade.

When we examine Merit’s network map in the context of the maps and graph presented earlier in this series of posts, it becomes clear that much of the network’s BTOP-supported expansion occurred in rural areas where the unused stocks of TVWS and EBS spectrum are relatively abundant.  For example, according to the map, Alpena in the northeastern lower peninsula, and Escanaba, in the southcentral upper peninsula, were both added to Merit’s fiber backbone network during this period.  And, as indicated in the spectrum graph, each of these communities have more than 325 MHz of unused TVWS and EBS spectrum available.

As I see it, this combination of: 1) high-capacity open-access nonprofit fiber and; 2) abundant spectrum unburdened by auction costs are the core network elements comprising the FAST (fiber and spectrum together) component of the FAST PIIPS connectivity model.

My next post will consider the PIIPS (public interest IP services) component of this model, with a particular focus on the institutional arrangements and motivations best suited to apply the model in rural areas facing challenging network economics.

Below are links to the other posts in this series. Feedback from readers, especially constructive criticism, is welcome.

• Part 1: Introduction
• Part 2: TV White Space as a Tool to Expand Rural Access
• Part 3: EBS Spectrum as a Tool to Bridge the Rural Homework Gap
• Part 5: The Role of Community Benefit-Focused Organizations
• Part 6: Research to Support Success

In the past I’ve written about the added value that public benefit-focused institutions (e.g., cooperatives, community anchor institutions (CAIs), municipal utilities, and social enterprises) can contribute to the connectivity equation, especially in relatively rural areas with high per-premise construction costs that make healthy competition among profit- and share price-driven entities extremely difficult if not impossible to achieve.

A key theme of this prior work has been that these public benefit-focused institutions tend to:

• have less demanding criteria regarding the level and timing of the direct financial return on their investment;
• are more likely to invest in directions they believe directly serve the public interest, as contrasted with organizations that prioritize private-owners’ interests but will, according to economic theory, serve the public interest if subject to market power-constraining competition, a situation unlikely to emerge in high-cost rural areas.

Other policy themes I’ve addressed in my earlier work relate to the development of the Internet and unlicensed spectrum models as key drivers of technical innovation, market entry and competitive options, and how these two developments relate to each other and can expand the range of entities and individuals able to cost-effectively use communication technology to pursue their goals, including CAIs, small businesses and others whose communication needs might otherwise be dependent on services provided by profit- and share price-focused companies with substantial and, in some cases, monopolistic market power.

New technical capabilities enable new models for serving the public interest

In 2008, when the FCC was weighing key policy decisions related to its treatment of TVWS, I drafted a policy analysis entitled Spectrum Policy 2.0: White Space, the Internet and the Public Interest.  In it I explained how the combination of an open high-speed Internet and expanded unlicensed spectrum opens the door to a more direct approach to serving the public interest than was possible when the 1934 Communication Act first established the electronic communication sector’s public interest standard and throughout the remaining pre-Internet era of the 20th century.  That paper introduced the term PIIP (Public Interest IP) to describe an Internet-era network and service model that provides an alternative to:

1. the dominant industry models that evolved over decades in highly siloed telecom and media markets, which were built around less capable technologies than are available today and have exhibited a strong tendency toward consolidation and anti-competitive market power and;
2. regulatory approaches applied to these markets intended to constrain the harms of excessive market power and to steer at least some impacts of private sector self-interested behaviors in the direction of a public interest standard that proved difficult to define and enforce, especially as technology and market dynamics evolved.

Later in 2008 I co-authored a report examining how the municipal fiber model is in some cases especially well suited to bring the full benefits of high-speed Internet access to insufficiently served local communities, especially if state legislatures don’t succumb to pressure from their state’s dominant ISPs (directly or via ALEC) to pass laws that impose unique and onerous burdens that weaken the economic viability of municipal networks, which these ISPs view as actual or potential competitors.  I’ve also found that electric cooperatives can be key players in extending fiber in very rural areas, and got involved in a particular approach to developing this potential around 2010, when USDA’s Rural Utilities Service (RUS) was investing stimulus funds to expand rural access.

More recently, I contrasted “generative” vs. “financially extractive” ownership and control models in both local access and backhaul (formerly Special Access, now Business Data Services) markets in a series of posts on the blog of Michigan State University’s Quello Center.  And most recently, I co-authored a Quello Center report entitled Wireless Innovation for Last Mile Access: An Analysis of Cases and Business Strategies that examined a range of emerging FAST PIIPS-friendly models for extending access to underserved populations in both rural and urban areas.

In addition to shedding light on the FAST (fiber and spectrum together) technology-related innovations that can help bridge challenging connectivity gaps, the wireless innovation report also considered the PIIPS element of the FAST PIIPS model.  It did so by examining the important roles played by public benefit-focused entities including: 1) nonprofit Research & Education Networks (RENs) that provide high-capacity fiber backhaul in many states, many of whose networks were expanded substantially with the help of stimulus funding; 2) community anchor institutions (CAIs), including schools, libraries and local governments, many of which are served by REN fiber connections and; 3) privately owned but public benefit-focused social enterprises, a good example of which is Axiom Technologies, an entrepreneurial ISP that, with the help of a Microsoft grant, has begun using TVWS to serve very rural and hard to reach parts of Maine.

Among the report’s conclusions was that:

“Given the economic and other challenges associated with bridging our nation’s remaining gaps in broadband access and the benefits it provides, locally-anchored enterprises and business models focused on sustainably addressing local needs may be better suited to this task than enterprises and business models focused on maximizing investor returns in national or global markets.”

It’s worth noting here that it is the usage of a broadband network rather than its mere existence that generates value and contributes to economic growth and quality of life. Certainly profit- and share price-maximizing access providers have strong incentive to attract paying customers and, in some cases, have invested in digital literacy programs coupled with discounted service and low cost devices to increase adoption.  But I’d argue that CAIs and social enterprises will, in most cases, place less emphasis than profit-focused firms on generating financial surpluses and more on supporting community benefits (an assertion I’ve suggested be tested by research focused on both access and backhaul networks).

Related to this is the fact that some CAIs, notably libraries and schools, bring to the table not only a strong motivation to promote network usage, but also a range of content and services that can help turn that usage into educational and other individual- and community-level benefits.  For example, according to a recent Pew Research survey, 78% of American adults (including 87% of millennials) believe public libraries “help them find information that is trustworthy and reliable,” and 76% of adults (85% of millennials) say libraries “help them learn new things.”

Michigan as a FAST PIIPS pioneer and testbed

As noted earlier in this series, Michigan has emerged as a leader in deploying networks that will help explore whether and under what circumstances the FAST PIIPS model can help expand broadband availability, adoption, usage and benefits in rural America. For example:

As discussed in a prior post, the EBS-based Educational Access Network (EAN) being deployed by Northern Michigan University–which uses LTE technology, relies heavily on backhaul provided by Merit Network, and involves cooperation with other schools and CAIs–represents an emerging education-focused FAST PIIPS model.

Michigan is also poised to become a pioneer in combining innovative use of TVWS with institutional innovation in ways that are consistent with the FAST PIIPS model.  As part of its 12-state Airband initiative (discussed in an earlier post), Microsoft, working with the Gigabit Libraries Network  (GLN), Merit Network, the Library of Michigan and other entities, is providing technical and financial support for TVWS deployments by: 1) three of Michigan’s public library systems (these projects are described briefly here and in somewhat more detail here) and; 2) Allband Communications, a fiber-based local access cooperative that operates in very rural areas of northeastern Michigan.

Working with Merit, Microsoft and the area’s community anchor institutions, Allband aims to use TVWS to cost-effectively reach still-unserved rural homes, businesses and CAIs beyond the reach of the fiber that it and Merit have deployed. Though the project’s use of available TVWS channels may be limited in its pilot phase, it seems unlikely to face a shortage of available TVWS spectrum since, as discussed in an earlier post, the rural areas it’s targeting enjoy a relative abundance of unlicensed FAST PIIPS-friendly spectrum.  In some areas this amounts to more than 200 MHz of TVWS, plus another 67.5-112.5 MHz in the EBS band, the use of which is being pioneered further north in the state by NMU and its partners.

Given that the key local participants in these Michigan EBS and TVWS connectivity projects are public universities and libraries working with cooperatively-owned fiber backhaul and access providers and a mix of other CAIs (e.g., schools, local governments, community centers), I consider them valuable testing grounds for the economic viability, impacts and evolution of FAST PIIPS connectivity models.

In the final post in this series I’m going to briefly outline a research agenda intended to help maximize the positive impacts on connectivity and digital empowerment of these pioneering projects and similar ones in other states, including those that will be part of Microsoft’s recently announced 12-state Airband initiative.

Below are links to the other posts in this series. Feedback from readers, especially constructive criticism, is welcome.

• Part 1: Introduction
• Part 2: TV White Space as a Tool to Expand Rural Access
• Part 3: EBS Spectrum as a Tool to Bridge the Rural Homework Gap
• Part 4: Fiber and Spectrum Together (FAST) for Rural Access
• Part 6: Research to Support Success

In announcing Microsoft’s Airband project, company president Brad Smith cited “a need for improved data collection regarding rural broadband coverage,” adding that:

“The FCC can help by accelerating its work to collect and report publicly on the state of broadband coverage in rural counties, thereby aiding policy makers and the private sector in making targeted investments.”

While I wholeheartedly agree with Smith, I’d expand on his recommendation along the following lines:

Matching unmet needs with available resources & economically sound strategies 

In addition to an FCC effort to accelerate its collection and public reporting of broadband availability data, I’d also recommend an expanded data integration and analysis effort that might involve not only the FCC, but also other governmental and private sector entities.  The goal of this effort would be to develop a user-friendly and expandable data and analysis platform that contains not only the FCC’s existing data on broadband coverage (or lack thereof) referenced by Smith, but also additional components that add to the platform’s value in developing broadband policies and action plans at local, state and national levels.  Among the additional GIS layers this platform could include are:

• the amount of available TVWS spectrum;
• the amount of unlicensed EBS spectrum;
• service areas not only of wired and wireless broadband providers but also of rural electric cooperatives and municipal utilities;
• locations of community anchor institutions (CAIs), including universities, libraries, K-12 schools and district offices, government facilities and, where possible, the status (i.e., speed and cost) of their connectivity (for an example of the latter, see this interactive map from Education Superhighway.)
• location of fiber backhaul lines and points of presence, including those connecting CAIs;
• vertical assets suitable for antenna siting;
• topographical features (e.g., contours, vegetation) impacting wireless coverage;
• FCC and Census data on residential broadband adoption levels (as distinguished from measures of broadband availability)
• population demographics and density data, and business/economic data available from the Census Bureau and other government agencies.

Given that most of this data is already compiled by one or more entities, the incremental tasks involved relate more to data integration than to data collection.  Though the former is not a trivial task area, it seems manageable in relation to its potential value.

Making the data platform user-friendly and expandable

To extract the most value from this integrated dataset, it should be readily accessible and usable by a range of stakeholders.  With this need in mind, and citing the value of data access platforms like Waze and Data.gov, Blair Levin and Larry Downes suggest in a Washington Post op-ed piece that the FCC “create an interactive broadband dashboard…designed so that anyone can build new tools to analyze particular state and local infrastructure.” They recommend that this dashboard be “continually updated with the most current information on broadband technologies, speeds, performance and coverage…[and] include new data sources and reporting methods.”

In keeping with this expansive vision of a toolkit to support the research and planning needs of a range of broadband stakeholders, I’d recommend that it also include a regularly updated version of the kind of detailed comparative network cost modeling done in 2009-2010 by CostQuest Associates as part of the National Broadband Plan exercise.  This cost-modeling component would not only update costs for the network options considered in the original National Broadband Plan analysis, it would also consider cost data for emerging network strategies, including the “fiber + wireless” solutions discussed in this series of FAST PIIPS blog posts and in my recent Wireless Innovation for Last Mile report.  In addition to drawing on cost data available from these innovative projects, it would also integrate updated cost data available from public and private sector sources, including the FCC, consultants like CostQuest, and the TVWS-focused study recently conducted by Microsoft and Boston Consulting Group and cited in Microsoft’s Airband announcement.

For the comparative cost modeling component to be most valuable, the data platform should be designed so that the broadband dashboard enables stakeholders to compare the costs of alternative network models for the particular geographic areas of interest to them, perhaps with the ability to adjust some of the model’s parameters to better reflect the real-world situations these stakeholders face.

Also helpful for policymakers and planners would be data on broadband pricing.  Though this data is inherently “messy” due to heavy use of and frequent changes in multiservice bundles and promotional offers, it is tracked by commercial firms.  For example, pricing data compiled by Telogical was used in preparing the National Broadband Plan, which recommended that:

“The FCC and the U.S. Bureau of Labor Statistics (BLS) should collect more detailed and accurate data on actual availability, penetration, prices, churn and bundles offered by broadband service…and should publish analyses of these data.”

Telogical broadband pricing data has also been used in studies conducted by consulting firms and academics.  In one such academic paper, entitled Measuring Broadband Internet Prices, authors Gabor Molnar, Scott Savage and Douglas Sicker, noting the limitations on available broadband pricing data and its analytical utility, recommended an “industry-wide collaborative effort” to develop an “ongoing, rigorous, nationwide study” of U.S. broadband pricing trends.

As I see it, the integrated set of data and analysis tools discussed above would help policymakers, network investors and other stakeholders develop more effective strategies and action plans at local, state and national levels.  It would do so by enhancing their understanding of:

• remaining gaps in broadband availability, performance, affordability and adoption;
• the extent to which fiber, spectrum and other network resources are available to help bridge these gaps;
• what types of public benefit-focused or shareholder-focused entities might be available to employ these resources to support the application of FAST PIIPS-friendly or other network models and;
• the relative costs associated with applying different network strategies in specific local communities.

Learning from Airband and other pioneering projects

In addition to the kind of GIS and financial model-intensive research described above, I’d also recommend an effort to update, expand and deepen the kind of case study research we did for the Wireless Innovation for Last Mile Access (WILMA) project.

Because our WILMA research included several groundbreaking network projects still in early stages of development, we were very limited in our ability to investigate the challenges these projects faced, how they responded to these challenges, the lessons they learned, and the impacts they achieved.  Given their significance as pioneers in the “last mile innovation” space, I strongly believe they deserve in-depth follow-up study.

Another valuable and timely direction for expanding the body of case study work in this space would be to examine the development of the Airband projects from a range of perspectives, including technology, business models, intra-and inter-organizational structure and dynamics, and economic and social impacts and outcomes.  With projects in twelve states, the Airband initiative–especially if carefully and impartially studied–can generate an abundance of lessons regarding strategies that can help address the nation’s remaining rural broadband availability gaps, including the role of TVWS in doing so.

How to proceed, and who will take the lead?

Given its resources and role, an important source of data and potential lead player in developing such a data platform is the FCC, perhaps working with other federal agencies that have datasets and areas of responsibility related to the availability and use of broadband connectivity (e.g., economic and job growth, effectiveness of education and healthcare sectors, etc.).  A variation on this would be for one or more states, private companies and/or foundations to take the lead, hopefully with active cooperation from federal agencies.

Personally, I’d love to see Michigan–where I’ve been working on broadband-related policy issues since returning here a few years ago–be involved in such a project.  As noted in earlier posts, the state has already emerged as a leader in deploying FAST PIIPS-friendly initiatives employing both EBS and TVWS spectrum.  Both of these Michigan initiatives involve active support from Merit Network, one of the nation’s oldest and largest Research and Education Networks (RENs), while the TVWS projects also involve Microsoft, the Library of Michigan (part of the state’s Department of Education), the Gigabit Libraries Network, and three of the state’s local and regional library systems.

The type of research and analysis capability I have in mind could also benefit from the expertise and years of Michigan-focused work of Connect Michigan and its parent Connected Nation.  In addition to many years of broadband-related GIS work in Michigan and other states, Connected Nation also operates other research-intensive programs, including a Connected Community Engagement Program that uses surveys and other tools to help local stakeholders develop, execute and monitor broadband-related Community Technology Action Plans. Additional expertise to support this kind of collaborative research initiative could come from Michigan’s universities. [Full disclosure: I worked with Michigan State University’s Quello Center from 2014 to 2016 on broadband related issues and have talked to Connect Michigan about helping them put together a series of case studies focused on innovative approaches to expanding broadband connectivity and digital empowerment.]

Below are links to the other posts in this series. Feedback from readers, especially constructive criticism, is welcome.

• Part 1: Introduction
• Part 2: TV White Space as a Tool to Expand Rural Access
• Part 3: EBS Spectrum as a Tool to Bridge the Rural Homework Gap
• Part 4: Fiber and Spectrum Together (FAST) for Rural Access
• Part 5: The Role of Community Benefit-Focused Organizations