During the past few days I’ve been: 1) reviewing the BTOP program evaluation study’s final report prepared in 2014 by ASR Analytics and; 2) doing some homework aimed at better understanding issues related to the FCC’s pending special access study and proceeding. The former relates to a pending Quello Center research project, while the latter was prompted by release of a paper written by Mark Cooper, Director of Research at the Consumer Federation of America (CFA). Cooper presented the paper, entitled “The Special Problem of Special Access: Consumer Overcharges and Telephone Company Excess Profits,” at an April 5 event sponsored by the New America Foundation.
In his paper, Cooper made a theoretical and empirical case that “large incumbent telephone companies have engaged in abusive pricing practices” for special access services that have resulted in economic harm exceeding $150 billion over the past five years:
Today, special access is a $40 billion per year business, which works out to about $300 per household, which is almost equal to what they spend on landline telephone service…
This paper shows that about half of the total bill paid to the large incumbent local phone companies for special access service, who control between five-sixths and nine-tenths of the special access market, is the result of the abuse of market power – i.e. setting prices far above costs to earn excess profits…
Because of the importance of special access as an intermediate good, the $20 billion in annual overcharges suppresses a significant amount of economic activity, reducing economic output by at least another $20 billion. The magnitude of the harm has been growing steadily, so that the cumulative value of economic losses over the past five years is in excess of $150 billion.
While there’s much to consider and debate in Cooper’s analysis and, more broadly, related to the FCC’s pending special access proceeding (something that may be done in the future by me and/or MSU colleagues on this blog), for now I’m only going to use Cooper’s general argument as a jumping off point to consider a few metrics that struck me in ASR’s BTOP evaluation study.
As explained in the introduction to its final report, published in September 2014, ASR Analytics summarized the scope of and general methodologies used in its study:
The scope of work includes an assessment of the benefits that BTOP grants are having on broadband availability and adoption, and in achieving social and economic benefits in areas served by the grantees…
ASR developed its conclusions based on a mixed- methods approach that includes comparative case studies of BTOP-funded projects, input-output analysis of the short-term economic impacts of all BTOP budgetary spending, and a matched-pairs analysis of the counties served by infrastructure grants in the evaluation study sample.
The ASR study had lots of ground to cover and, in my view, did an admirable job of using mixed methods to leverage available qualitative and quantitative data to evaluate a large number of projects spanning three major categories: Comprehensive Community Infrastructure (CCI), sometimes referred to as “Middle Mile Fiber” projects, Public Computer Centers (PCCs) and Sustainable Broadband Adoption (SBA) projects. And, importantly, as ASR noted in the introduction to its final report, it was “required to provide NTIA with all data that created a foundation for the [report’s] analysis and conclusions, as well as all data that could be utilized by future researchers.”
In this blog post I’m going to focus on a few elements of ASR’s analysis of the impacts of CCI projects, whose purpose and nature are, in my view, closely tied to the issues surrounding the special access market. This is because most if not all CCI projects involved construction of fiber networks providing high-capacity, high-reliability connectivity to “community anchor institutions” (CAIs), including schools, libraries and government and healthcare facilities, as well as providing non-discriminatory open-access backbone connections to wholesale and last mile service providers.
As I see it, the BTOP CCI projects represent a significant wave of new players entering the special access space, the bulk of whose construction costs (up to 80%) were covered by federal grants, and whose terms of service are required to meet certain nondiscriminatory “open access” requirements. And, conveniently and importantly, thanks to NTIA and ASR, these projects have been the subject of careful study from their inception.
Though there’s plenty we can learn from the ASR study of CCIs, I’m going to focus here on some high-level numbers that the study gathered, estimated and extrapolated.
According to Table 7 on pg. 15 of ASR’s final report, the total amount (including both federal grants and matching funds) budgeted for 109 CCI projects was $3.9 billion. The table also indicates that, at the time the study was done, these projects had connected 21,240 CAIs, at a budgeted cost of $184,141 per CAI. Assuming federal grants paid for 80% of this total cost, the average federal grant amount per CAI would be in the neighborhood of $147,300.
Table 13 on pg. 34 of the report shows the changes in subscription speeds and pricing experienced by the 86 CAI locations providing this information to ASR. The table shows very large increases in speed and, depending on the category of CAI, dramatic 94-96% average reductions in per-Mbps pricing. Table 14 on pg. 36 uses these reported changes in speed and price to extrapolate CAI cost savings from switching to CCI-provided fiber connections. Averaged across all CAI categories, the per-CAI annual savings amounted to $236,151.
So, in a single year, the average CAI saved well more ($236,151) in operating costs than the total capital cost ($184,141) required to connect it to a CCI fiber network.
These direct cost savings to CAIs were only part of the impacts considered by ASR. It also used previously developed models to estimate other economic benefits, as explained on pg. 33 of its final report:
Increased economic output: The largest long-term social or economic impact due to BTOP infrastructure spending is the yearly increase in GDP in the areas served by the new broadband infrastructure. ASR used two studies, Czernich et al. (2011) and LECG Ltd. (2009), to extrapolate the increase in economic output that could be expected in counties receiving BTOP- funded infrastructure. For the base case of a 2.0 percent increase in broadband availability, BTOP infrastructure spending could be expected to yield $5.7 to $21.0 billion in increased output annually using results from Czernich et al. (2011) and LECG Ltd. (2009) as the bases for extrapolation, respectively.
Long-term increased levels of employment: Kolko (2010) and Gillett et al. (2006) provide a basis for estimating the long-term increase in employment due to BTOP-funded infrastructure spending. Based on Kolko’s estimates, the additional broadband infrastructure provided by BTOP could be expected to create more than 22,000 long-term jobs and generate $1.1 billion in additional household income each year. Results from Gillett et al. (2006) suggest at least 6,900 long-term jobs could be created in the year following the construction of BTOP infrastructure, and potentially each year for at least the next four years due to increasing employment growth in areas with new broadband availability. These employment increases would result in a $328 million increase in household income for each year employment increases by the estimated amount in newly served areas.
Value to new subscribers: The Allen Consulting Group (2010) finds the value of broadband Internet access to the average American household is about 3.4 percent of average household income. Using the base case to determine the number of households adopting broadband, this translates into an estimated value of broadband to new subscribers of $2.6 billion per year.
Given all of the above, the CCI component of NTIA’s BTOP program strikes me as a very good investment of public funds in that it: 1) delivers substantial direct and indirect net social value, as suggested by the ASR study and; 2) helps correct the substantial excess-profit market failure suggested by both the CFA analysis and the dramatic cost savings reported by CAIs after connecting to BTOP-funded CCI fiber networks. And given these strong indicators of net social value, I’d suggest that the federal government consider expanding its CCI investment in geographic areas that the FCC’s special access data collection project indicates still face a lack of competitive options and an abundance of excess-profit-extracting prices in the special access market.
I’ll have more to say about this perspective in future posts.
After thinking and writing about the benefits of fiber-to-the-premise (FTTP) networks for a number of years, I’m happy to report that I’ve finally joined the small but growing population of U.S. households connected to the Internet via fiber networks delivering symmetrical gigabit speeds. I’m getting service from LightSpeed, a Lansing-based company deploying its network on a neighborhood-by-neighborhood basis (see this map for more details on the status of its network and service rollout).
The installation happened yesterday (the fiber drop was installed the day before). It went smoothly, with Jeremy and Chris connecting the drop to a fiber network interface in our basement and from there running CAT-5 cable to a router in our living room. As they worked, they patiently fielded the many questions I had and my long-winded comments about competitive dynamics, policy, etc.
Though my wife and I won’t be accessing the network on a regular basis until we move into our new house next week, I did have the pleasure of watching the Speedtest.net meter zoom up to the 1 Gbps range for both download and upload speeds at the end of the installation process. I also learned that while our fairly old laptops can reach this range when connected to the router via CAT-5 cable, their fairly ancient Network Interface Cards maxed out at around 100 Mbps in both directions when relying on Wi-Fi. But this was still by far the fastest speeds I’ve ever had in my home, especially in the upstream direction (the photo below shows Jeremy and Chris demonstrating speeds in the 500 Mbps range via a Wi-Fi connection on their newer-generation laptop, after hitting gigabit speeds via a hardwire connection to the router).
LightSpeed’s monthly price for symmetrical gigabit service is normally $70 (which I believe is also what Google charges in its Google Fiber cities). As an “early adopter,” however, I’m paying only $49 a month, a price that’s guaranteed for two years. Though this price no doubt takes a bite out of LightSpeed’s margin for its “early adopter” customers, it’s likely to help them capture market share as they build out their network. And in a high-fixed cost business like this, penetration is key to both short-term and long-term financial health.
One of the things that strikes me about LightSpeed is that it is not offering the “triple-play” bundle that has been the focus of cable/telco competition for many years. Instead, their value proposition is pretty simple: you get attractively-priced access to the Internet at blazingly fast symmetrical gigabit speeds (hopefully with the kind of reliability fiber can deliver but is much harder to match with the fiber-copper hybrid networks still mainly used by incumbent telcos and cablecos). You then have free rein to make (and change) the online services and devices you use to access the “voice” and “video” applications we once knew as “telephone” and “television.”
I’d love to do a survey of homes in areas served by LightSpeed to get a sense of who (and why and how many) are signing up for the company’s service, and what their experience is like. The obvious (and probably true) speculation would be that it’s mainly younger cohorts (e.g., Millennials), which research has shown account for a disproportionate share of the (cable and wireline telephone) cord-cutter population. That being said, I wonder how many older households, like my own, will jump at the chance to forego bundles loaded with things we don’t want (including extra fees for set-top rentals, etc.) and, instead, to simply get connected to a fast reliable network and then decide—-independent of the business arrangement we have with our network access provider—how best to use that network to support our needs and desires for communication, information, entertainment, commerce, etc.
I also wonder the extent to which the availability in a community of these “simple but super-fast” connectivity offerings will push incumbent providers away from their increasingly artificial “siloing-then-bundling” of services. There are already signs that this is taking place, though still on a modest and incremental scale.
My own view is that the longstanding focus on multiservice network operator-controlled bundles will continue to erode, especially to the extent that “simply super-fast access” options like LightSpeed expand their footprint. Among the other factors I see as supporting this shift are: the proliferation and increased penetration of online content and services, including Netflix, Hulu, YouTube and online-only options from longtime cable TV anchor service HBO; the increasing penetration and capabilities of smartphones, tablets and other access devices not (or at least less) controlled by network operators and; the ongoing shift in demographics favoring “fast affordable connectivity + apps of my choosing” vs. “a limited selection of bundled options offered by my network operator.”
Though not much can be predicted based on a sample of one, what I can say at this point is that I look forward to my own household’s transition down this road, which combines much faster and more reliable, symmetrical and affordable network access with expanded freedom of choice. And, as both a customer and industry observer, I hope that the increasingly healthy competitive environment emerging in the Lansing area remains so, and breeds increased innovation, investment and consumer choice.
I’ve been thinking a lot lately about the design and management of society’s core infrastructure systems (which I define broadly to include things like healthcare, education, housing and “money”) in an era marked by several important trends (for reasons suggested below, I refer to this as the “digital anthropocene”):
To flesh this out a bit, here are some examples of developments reflective of these trends (some of which I’ve already written about here):
From the Detroit Free Press (bolding is mine):
Businessman Dan Gilbert is not only buying dozens of buildings in downtown Detroit, he’s also linking them with some of the fastest Internet connections anywhere.
Gilbert, founder and chairman of Quicken Loans and Rock Ventures, said this weekend he has formed a new “community investment initiative” called Rocket Fiber LLC to provide faster Internet connections in downtown Detroit…Construction is already underway…in the greater downtown area. Eventually the goal is to expand the service to other parts of the city.
Even if Gilbert’s venture decides it’s not economical to extend fiber to every (or even most) locations in the city (a decision that strikes me as very likely), it might be willing to extend high-capacity fiber trunks around the city to feed wireless access links that use unlicensed spectrum to provide affordable high speed connectivity to more of the city’s homes and businesses.
There are already several grassroots community WiFi networks being deployed in neighborhoods around the city by a group called Detroit Digital Stewards, under the Allied Media Projects (AMP) umbrella.
If the price of connecting to Rocket’s fiber trunk lines is low enough, this model could support inexpensive high-speed connections deployed neighborhood by neighborhood by groups like Digital Stewards, which isn’t just focused on connectivity, but also on developing applications that support community development.
It will be interesting—and perhaps important for Detroit’s revitalization efforts—if “downtown-focused” projects like Rocket Fiber can develop strong cooperative relationships with “neighborhood-focused” initiatives like AMP and its Digital Stewards program. And perhaps the task of deploying affordable broadband Internet infrastructure—which can support a wide range of revitalization-focused collaborative activities—is a good place to start building such “downtown-neighborhood” alliances.
The Motor City could certainly use a boost in connectivity. According to an analysis of Census Bureau data, Detroit has one of the nation’s lowest rates of Internet access (see tables below).