Below are the comments of the National Digital Inclusion Alliance regarding HUD’s proposed rule in Docket No. FR 5891-P-01, “Modernizing HUD’s Consolidated Planning Process To Narrow the Digital Divide and Increase Resilience to Natural Hazards”. Here is a PDF of our consolidated planning comments. We also submitted comments regarding HUD’s proposed rule on the Installation of Broadband Infrastructure. Since infrastructure is not our expertise, for those comments we referenced and supported the comments of the City of Seattle and the National Housing Conference. Here is the PDF of our infrastructure comments.

The National Digital Inclusion Alliance (NDIA) respectfully submits these comments on HUD’s proposed rule in Docket No. FR 5891-P-01, “Modernizing HUD’s Consolidated Planning Process To Narrow the Digital Divide and Increase Resilience to Natural Hazards”.

NDIA represents leaders of community organizations, public libraries, local governments, housing authorities and other institutions working in our communities to reduce digital disparities among our neighbors. Our affiliates help millions of disadvantaged Americans to join the 21st century mainstream through digital literacy training, public Internet access, affordable home broadband services, and digital inclusion advocacy.

On behalf of this unique network of experienced community leaders and practitioners, NDIA advocates digital inclusion policies based in our shared conviction that broadband adoption is most effectively promoted by collaborative, community-driven efforts that combine affordable home broadband service, public broadband access, and technology training and support provide by trusted community partners.

NDIA currently counts 193 affiliated organizations. Our affiliates currently include 33 national nonprofits and 139 local public and nonprofit organizations in 31 states and the District of Columbia, and in 41 urban metropolitan areas. Local NDIA affiliates include fourteen municipal governments, three housing authorities, twenty-two local public libraries and regional library councils, and eighty-one local nonprofit organizations. A full list and map of NDIA affiliates can be found at http://www.digitalinclusionalliance.org/members/.

The National Digital Inclusion Alliance has strongly supported the ongoing effort of President Obama, the Broadband Opportunity Council, and HUD to foster broadband deployment and adoption for all Americans, including the 45% of American households with incomes below $35,000 a year which still lack home broadband access. We recognize that the proposed rule is an important part of this overall effort.

We welcome the opportunity to offer the following comments on proposed changes in HUD’s rules governing the Consolidated Plans of state and local governments which are addressed to the broadband needs of low and moderate income residents.

1. Consultation and citizen participation requirements (§§ 91.100.91.105. 91.110, 91.115).

The proposed rule would amend these requirements to specify that local governments and States must consult with public and private organizations, including broadband Internet service providers, and other organizations engaged in narrowing the digital divide.

Further, the citizen participation plan must encourage their participation in implementing any components of the plan designed to narrow the digital divide for low-income residents.

NDIA supports and welcomes these changes. Local organizations that provide training and assistance in obtaining and using broadband are critical to effectively closing the digital divide. As we will suggest below (section 3a), they may also be critical sources of local information for Housing Market Analyses, especially regarding the scope and causes of broadband non-adoption by community residents.

We suggest adding language to clarify the range of “organizations engaged in narrowing the digital divide” to include “local social service and public agencies providing digital literacy, public internet access, or other broadband adoption programs. These may include, but are not limited to adult literacy and education providers, K-20 schools and youth program providers, libraries, and small business and workforce training program providers”.

We also recommend that the term “citizen participation plan” in this context be changed to encourage “resident participation” to ensure inclusion of technology disadvantaged legal residents who may not yet have citizenship.

2. Contents of Consolidated Plan (§§ 91.5, 91.200, 9.200, 91.210, 91.300, 91.310).

First, the proposed rule would require that, in describing their consultation efforts, local governments and States describe their consultations with public and private organizations, including broadband Internet service providers, [and] other organizations engaged in narrowing the digital divide…

Second, the jurisdiction must also describe [as part of the housing market analysis in its

consolidated plan] broadband needs in housing occupied by low- and moderate-income households based on an analysis of data for its low- and moderate-income neighborhoods in the National Broadband Map. The National Broadband Map Web site may be accessed at http://www.broadbandmap.gov/. Grantees may also use broadband availability data in the FCC Form 477 or other data identified by the jurisdiction, for which the source is cited in the jurisdiction’s Consolidated Plan. These needs include the need for broadband wiring and for connection to the broadband service in the household units, [and] the need for increased competition by having more than one broadband Internet service provider serve the jurisdiction.

NDIA strongly supports the general purpose of this proposed change, but believe the data types and sources suggested are too limited for actionable analysis, either of the three specified categories of “need”, or of the market realities, community impacts and opportunities they represent.

1) The National Broadband Map and FCC Form 477 broadband availability data. The National Broadband Map, which was based on data provided by state broadband offices (but ultimately by providers), is no longer being updated due to lack of funding. Similar broadband availability data (maximum download and upload speeds by Census block by provider by technology), reported by providers to the FCC via Form 477, will continue to be updated on a semiannual basis, but is currently provided to the public only in the form of large state CSV files, and only for fixed technologies. We understand there are plans to make this data more accessible to communities via a mapping tool or other application; this should increase its value for the purpose described in the proposed rule change. But given the rapidity of change in this market and the increased prominence of wireless technologies in some traditional wireline providers’ basic service offerings, HUD should be cautious about enshrining either of these sources as the default authority for grantees’ use in Consolidated Plans that will be developed two, three, or four years from now.

That said, at this point in time, the National Broadband Map and Form 477 data in combination do offer states and communities a convenient and reasonably reliable way to determine a) whether specific planning areas, e.g. city neighborhoods, have access to broadband services, and b) the identities of providers offering that access and their technologies.

But it’s important to recognize that neither source’s deployment data, based mostly or entirely on ISP self-reporting, carries any real guarantee of reliability. And because the reporting instrument asks only for the ISP’s fastest download speed for each of its technology offerings, Form 477 data will seldom, if ever, enable local planners to determine the quality of broadband services available at specific locations – or even the majority of locations – within a Census block. (Within a group of four inner-city Census blocks in Cleveland, for example, an NDIA affiliate recently found that residential download speeds offered by the AT&T website “availability” tool for specific addresses ranged from “no service” to 6 mbps, even though both the Broadband Map and most recent Form 477 data indicate maximum available download speeds between 1.5 mbps (VDSL) and 3 mbps (asynchronous xDSL). A nearby block, for which AT&T’s 477 data shows a VDSL download speed of 18 mbps, seems to have actual offered speeds for almost all addresses of only 6 mbps down; the exceptions our affiliate could find were four homes on the edge of the block adjacent to a VRAD installation, for which the website tool shows available Internet speeds much greater than 6 mbps, but only when bundled with video service.)

So while either of these data sources will help planners to determine how many ISPs offer residential Internet services in a particular Census block, neither is a source of reliable information regarding the speeds or prices at which those services are actually being offered to most households – which of course is critical to judging whether those services are truly competitive.

2) Sources for actual household broadband connection. The proposed rule offers no suggested sources for states and communities to assess the extent to which “the need… for connection to the broadband service in the household units” is being met – i.e. the extent to which broadband services have actually penetrated the market of low-to-moderate income households in a given community. For many low-moderate communities this will be a far more important starting point for a public discussion of digital access and equity than the nominal “availability” of broadband service.

There are two readily available Federal sources for actual household connection data which should be specified, at least as options, by the rule.

FCC Form 477 Census Tract Data on Internet Access Services. The first source is a separate set of data collected by the FCC via the semi-annual Form 477 ISP reporting process, and made available for download on the FCC website.2 Each provider must report the number of fixed residential Internet connections above a specified speed threshold that it provides, as of the reporting date, within each Census tract. The FCC sums these reported connections for each tract and compares the total to the tract’s total Census households. The tract is assigned a code (0 through 5) indicating its number of connections per thousand households, with “1” indicating 1 through 199, “2” indicating 200 through 399, and so on. These codes are available for download and make it reasonably easy to rank or map a state or community’s Census tracts by household broadband penetration. The FCC also provides a national interactive color-coded map of these rankings on its website, which, while fairly high-level, does offer both states and local communities an easy first look at their tracts’ codes.

American Community Survey. The second source is the Census’ American Community Survey data on household computer ownership and Internet access, provided in Tables B28002 through B28009. These tables first appeared in the 2013 ACS and are currently included only in the 2013 and 2014 One Year Estimates, which means that they’re available only for “places” with populations of 65,000 or more – not for tracts – which makes them of limited value for local Consolidated Plans. But in the normal course of events these tables will become part of the Five Year Estimate – and thus available for Census tracts as well as smaller communities – in the 2017 ACS published in 2018. So detailed Internet access data from the American Community Survey, including household demographics, should be very useful for Consolidated Plans filed in 2019 and beyond.

Local technology adoption surveys, ISP reports on uptake of low-income resident broadband offerings, and crowd-sourced surveys of actual speeds are also potential sources of this data which local planners should be encouraged to consider.

3) Making data on broadband availability and market penetration actionable by low to moderate income communities.

a) Looking at causes of digital exclusion other than availability. The Consolidated Plan process “serves as the framework for a community-wide dialogue to identify housing and community development priorities that align and focus funding from HUD’s formula block grant programs.” To fulfill this purpose, the broadband access assessment proposed for inclusion in the Housing Market Analysis must go beyond the nominal availability of infrastructure, the competitiveness of the local market and even the actual degree of household Internet penetration, as important as these “needs” are.

Millions of low-income households without home Internet access live in communities, including big-city neighborhoods, where local cable and telecom companies have offered competing high-speed Internet services for a decade or more. In recent years these fixed broadband providers have been joined in the market by 4G providers. Yet household Internet connection rates among poor residents in these communities often remain below 50%. (See http://www.digitalinclusionalliance.org/blog/2015/9/20/worst-connected-cities-2014)

If simple availability and provider competition were enough to get these households across the digital divide, it would have happened long ago. National survey research has clearly established that factors other than nominal availability – including cost, fear, and lack of technological literacy – play major roles in deterring broadband adoption, especially by poorer, less educated and older Americans.

In creating a framework through its Consolidated Plan process for community dialogue leading to possible action toward greater digital access and inclusion, HUD should explicitly encourage governments, which find low rates of household Internet access among their low-moderate income residents, to investigate possible causes other than physical availability of service, including:

 unaffordability of available Internet services to low-income residents;

 a lack of convenient opportunities for residents to gain digital literacy skills, including community training programs;

 a failure to communicate the value of available Internet services and tools (in education, health, employment, commerce, civic participation, etc.) to unconnected residents and communities;

 enrollment or materials design, lack of awareness or other barriers that limits adoption of ISP provided low income internet programs;

 other factors specific to communities, such as language, cultural barriers, etc.

In most cases useful information in these areas will come, not from easily available public data, but from the people included and questions asked in consultations and community dialogues – and possibly from communities’ own local research initiatives. Whatever the sources may be, governments should be asked to describe what they learn about underlying causes of persistent digital exclusion of low-moderate residents (to the extent the data shows it to be a community problem) in their Housing Market Analyses.

This should make community dialogue regarding possible investment in digital inclusion measures and programs a great deal easier and more productive.

b) Addressing obstacles to community action to increase Internet access. There is little reason for HUD to require governments to undertake the extra work entailed by the proposed rule changes, unless HUD hopes and expects that some of those governments will eventually decide to invest CDBG, HOME and other program resources in efforts to narrow the digital divide in their communities.

While not technically within the scope of public comment in this rulemaking proceeding, NDIA would like to recommend a companion effort by HUD to assess how existing rules and legislation governing these programs may limit the ability of grantee governments to make that decision.

One example is the “public services cap” on grantees’ permissible use of Community Development Block Grant dollars. Any local investment of CDBG funds in digital literacy training, technical assistance or even consumer premises equipment to support household Internet adoption is currently classified as a public service expenditure and limited by the cap, which means it must compete for a fixed pool of dollars with all kinds of ongoing community needs like senior services or safety organizing. To encourage creative, effective community digital inclusion initiatives, HUD should consider ways to reduce or eliminate this obstacle.

Another example is the absence of HUD guidance regarding the status of broadband as an eligible expenditure for affordable rental housing. HUD should consider adopting policies that treat broadband as standard operating cost and eligible expenditure in affordable housing. This recognition of the utility of broadband services would enable localized programs that could reduce cost and barriers to adoption by aggregating costs and internet delivery. This could further enable models developed in conjunction with use of the upcoming FCC Lifeline broadband program for low income residents.

c) Supporting best practices to narrow the digital divide in low and moderate income communities. HUD programs such as ConnectHome and Neighborhood Networks have been shown to be important targeted efforts to foster broadband access and adoption. HUD funding for national coordination and development of best practices, along with assistance to local implementation, would be an important complement to the proposed changes in the Consolidated Plan process, helping to ensure that HUD and its partners can meet both national and local goals for broadband access and adoption.

Thank you for the opportunity to offer these comments on behalf of NDIA’s affiliates.