Building for the Future Instead of Catching Up: A Proposal for a National Connectivity Trust

For decades, U.S. telecommunications policy has struggled to keep pace with the transition from analog voice telephone service to digital, Internet Protocol-based advanced telecommunications.

Ideally, better policy would have been put in place in the early 1990s to build what then Vice President Al Gore envisioned as an "information superhighway," reaching nearly every American home, school and business just as Plain Old Telephone Service (POTS) did for the latter half of the 20th century.

Over the past three decades, policymakers and regulators have implemented various subsidy programs to support the transition from POTS copper networks to fiber optic connections. Instead of building for the future, they have tried to catch up, and they have come up short and late.

I've drafted a framework for a different approach, a single institution I'm calling the National Connectivity Trust. The full discussion draft is available here [link to PDF]. What follows is a summary.

Why the current approach isn't working

The main mechanism, the Universal Service Fund (USF), is funded by a charge on interstate voice revenue, a market in long decline. In the fourth quarter of 2026, that charge will reach a record 42 percent, the second record quarter in a row. Loading ever more of the cost of future networks onto a shrinking base of voice calls can't last.

Meanwhile, federal support is spread across more than 133 programs administered by 15 agencies. GAO describes the result as "fragmented and overlapping". States run their own programs on top of that.

Even after BEAD, the largest construction program yet, about 1.1 million locations remain unfunded. The figure rises to 1.6 million if satellite awards aren't counted as service.

Construction grants also leave a gap: they pay for building networks, not for running them. NTIA now requires BEAD recipients to certify that they won't seek future operating subsidies. In high-cost rural areas, operating costs can exceed subscriber revenue even when construction was fully paid for.

Costs keep rising. In 2025, median fiber construction costs reached $18 per foot underground and $8 per foot aerial, and 88 percent of builders expect further increases in 2026.

Then there is affordability. The Affordable Connectivity Program was serving 23 million households when its one-time funding ran out. Doug Dawson of CCG Consulting reports a rise in customer losses ISPs now see comes from households that can't afford the monthly bill. Smartphone-only households grew from 7.5 percent in 2017 to 11.5 percent in 2024.

A different approach: finance and operate for the long term

The National Connectivity Trust would be a congressionally chartered nonprofit federal corporation, tax-exempt under IRC § 501(c)(1) and administered through nine regional offices plus a national Tribal Office. It would take over every federal program that pays for networks or service, ending the grant and subsidy administration roles of the FCC, NTIA and USDA's Rural Utilities Service. The FCC would keep its regulatory work: pole attachments, spectrum, numbering and consumer protection.

The Trust would do four things.

1. Lend rather than grant. Its core is a long-term, low-interest loan facility for middle-mile and last-mile fiber, with priority for consumer cooperatives, electric cooperatives, and municipal and Tribal sponsors. Nonprofit and cooperative operators need no return on equity and pay no income tax. A 2 percent, 30-year loan to such a sponsor supports roughly twice the construction cost that a private operator's 9 percent return target would. Replacing aging poles, a major obstacle to aerial fiber in rural and exurban areas, is an eligible cost.

Where a project can't repay all its costs, part of the loan is forgivable. The share forgiven is set by a published national cost model, and sponsors who request less get priority. Forgiveness is earned in stages:

• 25 percent at construction completion.
• 25 percent once service is established.
• 50 percent spread over ten years of compliant operation.

Sponsors who fail to perform owe the balance. This keeps the focus on building and operating networks at affordable rates, not on competing for free money.

2. Use satellite where fiber can't pay for itself. In very-low-density and remote areas, the Trust would buy satellite service competitively, reviewing every five years whether those areas can move to fiber. Satellite has limits. As Blair Levin, former policy advisor to New Street Research, put it, satellite "can provide an important connection where better options are not available" but is not "interchangeable with fiber for the long-term demands of an AI-driven economy". The five-year review keeps satellite a bridge to fiber rather than a permanent substitute.

3. Make service affordable everywhere. Trust-financed fiber networks would offer one residential rate, $65 a month for symmetrical 1 Gbps service, with no speed tiers. Business service would have speed tiers. Customers could pause and restart service without penalty.

A permanent Affordability Credit of about $30 a month would replace the ACP, with no separate application. Every household in a high-poverty census tract would qualify automatically, as would households enrolled in SNAP, Medicaid and similar programs. The credit would apply to fixed service from any provider, including in urban areas where mobile-only use is growing.

4. Drop the map and the challenge process. There would be no national availability map, no "served/underserved/unserved" categories and no process for incumbents to challenge eligibility. Sponsors define their own service areas and must serve every location in them. Eligibility rests on financial viability, not on the speed of existing service. Some overlap with existing fiber, which incumbents often build only in their most profitable neighborhoods, is expected, but no forgivable money goes to duplicate fiber.

Who pays

The Trust would be funded by a Universal Connectivity Excise Tax on large edge providers and AI hyperscalers, the companies whose business depends most on these networks. The USF charge on voice revenue would phase out over three years.

U.S. digital advertising revenue alone was $294.6 billion in 2025. By my estimates:

• Building the networks would take less than 1 percent of that revenue.
• Adding the existing USF programs would raise the rate to about 3.4 percent.
• Adding the Affordability Credit would bring it to about 6.1 percent.

Does it pencil out?

Financial modeling suggests it can:

• Fully repayable loans work in exurban and rural aerial areas for cooperative-cost operators at the $65 rate.
• Mixed projects that combine cheaper and costlier areas can often repay in full.
• Lower-density areas need a forgivable share of roughly a quarter to two-thirds of construction cost.

Because satellite serves the most remote locations, the federal government would not need to own networks. It would take one over only as a remedy when a sponsor fails.

Federal law already says advanced telecommunications should reach "all regions of the Nation" at "reasonably comparable" rates. What's missing is an institution built to deliver that: sustainable funding, finance that doesn't depend on subsidy forever, and affordability built in rather than added as an afterthought. The National Connectivity Trust is my proposal for that institution. I welcome your comments.

Full discussion draft: National Connectivity Trust: A Framework for the Universal Connectivity Act (September 2026) [link to PDF].

The magical economic thinking at the center of the 1996 Telecom Act: Why Al Gore’s Information Superhighway wasn’t built

The Telecommunications Act of 1996 is now 30 years old, and there has been a lot of events, hearings and webinars, including a congressional hearing, an FCC series of panels, Public Knowledge, Benton Foundation, TPI, Brookings, Broadband Breakfast, and a few others, all easily findable on the web.

The Act was supposed to open the wired networks to direct competition, that would lower prices and bring in new and innovative services that would be available via a new fiber optic wire to the home and business. And it would be delivered to everyone, equally, as this Act was an update of the original Communications Act of 1934.

https://kushnickbruce.medium.com/telecom-act-is-30-500-billion-overcharging-the-digital-divide-and-delete3-by-fcc-chairman-carr-e50d0ab5940f

With these words, Bruce Kushnick, a longtime critic of U.S. telecom policy, sums up the flawed magical thinking that made the envisioned future state immediately preceding this regulatory overhaul impossible to attain. As Kushnick describes it:

Starting in the 1990’s, a few years before the Telecom Act, America was promised a new shiny fiber optic future. Seven holding companies had been created in 1984 and given control over the existing state telecommunications public utilities, which were based on copper wire.

And in 1992, Vice President Al Gore laid out the ‘Information Superhighway’, a fiber replacement of this existing copper wires.

The root cause is negligent policymaking. A dispassionate economic assessment would have determined advanced telecommunications like basic telecommunications before it that ran on copper networks providing analog voice telephone service functions as a natural monopoly like other utilities.

As such, facilities-based market competition would be highly unlikely to appear since the barriers to competitor entry against established incumbents would be too high and the future return on investment too distant.

As AT&T explained, the cost of modernizing the copper cable distribution network to fiber to the premise infrastructure is enormous. Its shareholders or investors and creditors and those of any other privately held company would be unable to bear them. Consequently, as Kushnick’s analysis alludes, multiple state level plans by the RBOCs to replace the copper with fiber were abandoned. The economics would not pencil out.

That reality would have been reasonably if not painfully obvious for any serious policymaking effort in the early 1990s, calling for a different approach: the one not chosen.

The seven holding companies to which Kushnick refers are the Regional Bell Operating Companies (RBOCs) formed out of the 1984 settlement of the federal government’s antitrust prosecution of AT&T in the 1970s. Had the proper policy been put in place, they could have formed the structural framework for Al Gore’s Information Superhighway that could have reached most every American doorstep by 2010.

How would it work? By creating a federal-private partnership with the RBOCs. The federal government would form a telecom authority that would take a majority stake in each of the RBOCs with an option for the RBOCs to negotiate a purchase of those stakes once universal fiber is built out in their respective regions as determined by a federal review.

This could have fostered a mindset of bandwidth abundance rather than scarcity, thus avoiding decades of “broadband” battles at all levels of government over a constantly changing definition of broadband, where is available and at what price. And decades of unproductive protectionist lobbying by incumbent telephone and cable companies to stymie publicly owned and operated fiber delivery infrastructure despite its lower cost structure.

Futurist Alvin Toffler foresaw ubiquitous fiber to the home -- in 1990

Like phones and VCRs, faxes will begin to appear in even the humblest homes, driven by the Law of Ubiquity. And so will fiber optic cables and other advanced technologies, whether paid for by the individual, the public, or by other users whose fees will subsidize service to those who can’t afford it.

The widest diffusion of communication capabilities is an inseparable part of the new system of wealth creation. The direction is almost inevitably toward what the old Bell phone company called “universal service”—i.e., ubiquity—combined with interactivity, mobility, convertibility, and connectibility.

Toffler, Alvin. Powershift: Knowledge, Wealth, and Power at the Edge of the 21st Century (p. 356). Random House Publishing Group. Kindle Edition.  

Not too late to build Al Gore’s “information superhighway.”

If Congress could turn back the calendar to the 1990s when then Vice President Al Gore envisioned a digital “information superhighway,” it could have adopted a different strategy than unsustainably relying on legacy voice telephone service to chip away at its construction and the unrealistic idea that investor-owned service providers would compete to build high cost, long ROI fiber telecommunications infrastructure to replace legacy metallic voice telephone and cable TV plant. Congress should have instead formed a Digital Infrastructure Authority.

The regionally administered authority would similar to the Federal Highway Fund fuel tax that operates as an excise tax on companies that originate and send traffic over this fiber optic freeway, something that has been proposed by service providers and their policy advocates. 

It would provide long term, low interest loans to publicly and consumer utility cooperative owned networks so as to not favor any privately owned provider and function as an open access network – a public resource available to anyone who wants to use it. The authority would also have the ability to purchase existing networks and their rights of way in order to increase economies of scale, cost efficient construction and network reliability.

There would be plenty of economic opportunity for private sector players to design, build and operate the network. And it would reach farther than fiber networks they could afford to build, which limit them to proscribed builds that favor densely developed areas but leave other areas without service.

The Digital infrastructure Authority would also make access more affordable for end users since it would not have to generate profits for investors or pay business taxes. By charter, it would reach any doorstep on the electrical grid.

The good news is it’s not too late four decades later for Congress to choose this course. Call it a course correction.

U.S. telecom policy split: broadbanders versus infrastructuralists

Lacking a global policy to support the modernization of legacy copper analog telephone infrastructure to fiber to support modern Internet protocol-based voice, video and data services, the United States has defaulted to multiple, one off piecemeal subsidy programs. Determining how these programs operate has resulted in further fragmentation into two camps. They are the broadbanders and the infrastructuralists.

The broadbanders have held sway since the 1996 Telecom Act. It directed the Federal Communications (FCC) to annually survey the deployment of advanced telecommunications infrastructure and identify and correct impediments.

The FCC chose broadband speed – how fast bytes travel to and from end user premises – as the metric by which to gauge deployment. As long as the numbers were increasing over the past three decades, the FCC declared sufficient progress.

Various subsidy programs also adopted this metric as an eligibility factor. Only premises offered specified broadband speeds falling below an arbitrary cutoff were deemed eligible. That in turn led to the creation of “broadband maps” to determine which addresses were considered served and thus ineligible and which were “unserved” or “underserved” and thus eligible.

The infrastructuralists argue this is inherently wasteful and short lived since what is deemed adequate “broadband speed” is dynamic and growing rapidly such that by the time subsidies are awarded, projects face imminent obsolescence. They favor subsidizing fiber to the premise delivery infrastructure because of its long-term life, relatively low upgrade costs and its capacity to easily accommodate the longstanding trend of increased device and data use.

The infrastructuralists’ influence peaked in 2021 when the Biden administration’s infrastructure legislation proposed appropriating $90 billion to subsidize public and utility cooperative owned fiber that the president noted have a lower cost structure since they don’t have to produce profit for investors.

Naturally, investor owned providers that dominate America’s market-based telecommunications were opposed. The bill was quickly scaled down and amended to favor the broadbander camp, using the broadband speed metric and related broadband mapping to determine subsidy eligibility.

The Biden administration proposed guidance for the amended measure’s Broadband Equity Access and Deployment Program that allowed states to prioritize fiber projects in parceling out grants awarded to the states under the legislation. That gave the infrastructuralists leverage.

But the broadbanders -- particularly wireless and low earth orbit satellite services – claimed that was unfair. Americans in areas with obsolete legacy metallic infrastructure needed better service decades ago and suffered long enough. We can provide it much faster than building out fiber to them "to get people online," they claimed, urging the feds to liberalize BEAD so some of the subsidies flow our way.

The broadbanders gained influence in the waning months of the Biden administration and now hold sway in the current Trump administration. In the BEAD battle, some states are claiming they know what’s best to meet the needs of their residents and businesses and insist fiber is the best use of taxpayer dollars. But the billions needed to build it largely come from Washington, giving federal policymakers the ultimate say.

As they have for decades, the broadbanders remain dominant over the infrastructuralists.

Quantum enabled SDN

The company operates over a million miles of fiber and cable that deliver internet to 31.5 million homes and businesses. That means that for any piece of data to travel from Point A to Point B, there’s a near infinite number of combinations, said Elad Nafshi, Comcast’s chief network officer. 

Data going into New York for example, could travel along the George Washington Bridge, the Lincoln Tunnel or the Amtrak train tracks, he said. And the fastest route also depends on other factors like, if there’s a fiber cut somewhere or a big surge of data into New York while everyone streams the Giants game. 

Being able to calculate, in real time, all those variables to determine the optimal flow of data and deliver it at the fastest speeds for the highest number of people is something conventional computers struggle with. But, “that’s something that Quantum could do extremely well,” Nafshi said. 

 https://www.wsj.com/articles/heres-how-quantum-computing-could-change-the-world-c7a995b1

This recalls the Cold War origins of the Internet. It was designed as ARPANET in the 1960s by the U.S. Department of Defense to provide a computerized governmental communication network that was self healing -- meaning it could automatically route around metro areas destroyed by nuclear weapon attacks.

Here, it is routing around network congestion. 

Playing the long game in the fiber gold rush: Large investor owned provider overbuilding publicly owned network in small town that didn't initially pencil for private investment.

Fidium’s arrival has flummoxed local officials, partly because its parent company, Consolidated Communications, declined to build broadband infrastructure in Arrowsic years earlier, said Don Hudson, another commissioner. “It came as some surprise when all of a sudden we started seeing, essentially, a duplicated system being built on top of ours,” he said. “If it wasn’t actually happening, it would be laughable.”

Consolidated Communications, Fidium’s parent, owns many of the telephone poles in town. The local group had to pay tens of thousands of dollars to put its cables on them. That ownership has made it simple for Fidium to begin installing its own fiber without any approval from Arrowsic officials.

Machias-based internet service provider Axiom Technologies runs the town’s broadband service. Its CEO, Mark Ouellette, was also surprised to hear of Fidium’s entrance into Arrowsic. His company, which provides internet to several Maine towns, isn’t backing down.

“Typically, when another provider is in the community with fiberoptics, it’s quite a challenging business case to be made to build out fiber on top of fiber already there, especially in small places,” he said. “We are going to make a strong case that community cyber connectivity is an important asset for the community … because we return a percentage of our fees to the customer, back to the town.”

https://themainemonitor.org/arrowsic-broadband-challenge-fidium-fiber-network/?ref=broadbandbreakfast.com 

This is an interesting situation. A bigger for profit fiber ISP is overbuilding publicly owned fiber distribution infrastructure and taking advantage of its reduced barrier to entry with its ownership of the pole distribution infrastructure. 

Consolidated Communications is likely playing the long game. A premises fiber connection has long term value given the 30-50 year life of the fiber plant. Large investor owned providers enjoy greater economies of scale and ability to spread costs than publicly owned networks serving a single small town such as the case here  -- where both public and privately financed fiber desire to capture and keep end users.

This is the dynamic driving the fiber gold rush. A decade or two earlier, the business case for investor owned fiber turned on short term ROI that disfavored investment in smaller, less densely populated areas like Arrowsic, Maine. Now it recognizes the long term value of owning the fiber premise connection as well as the potential to sell it to an even larger player with a similar outlook. 

Lutnick’s right. Americans aren’t getting the benefit of the bargain -- of universal service.

U.S. Commerce Secretary Howard Lutnick announced June 6 that the Trump administration would be revising the program rules for the $43.45 billion Broadband Equity and Deployment (BEAD) program, authorized by the Infrastructure Investment and Jobs Act of 2021 (IIJA).

Lutnick said the program requires retooling in order to ensure Americans can obtain the full “benefit of the bargain” Congress intended in enacting the IIJA: broadband deployment. The Biden administration, in keeping with the infrastructure construction and modernization intent of the IIJA, administered BEAD with an infrastructure focus and specifically fiber to the premises (FTTP) and middle mile advanced telecommunications infrastructure.

The current administration however is reverting to the policy framework in place since the 1990s. It defines “broadband” as a service based on specified “high speed” throughput. The infrastructure to deliver it isn’t specified in this “technology neural” policy. In the original version of the IIJA, it was: FTTP. That fell away in a subsequent amendment of the legislation. (See earlier blog post here).

By deemphasizing landline infrastructure and instead making BEAD subsidies available for cheaper and less reliable non-landline infrastructure delivered service as fixed wireless and low earth orbit satellite, the Trump administration will “connect more Americans to broadband more quickly, and at a lower cost to the American taxpayer,” said Lutnick, who also serves as acting administrator of the National Telecommunications and Information Administration (NTIA), charged with implementing BEAD.

However, Americans have never gotten the real benefit of the bargain: universal service of Internet protocol-based advanced telecommunications delivered by landline like voice telephone service before it. The expectation of that bargain was expressed as public policy in the Telecommunications Act of 1996.

According to the Federal Communications Commission, the Act “expanded the traditional goal of universal service to include increased access to both telecommunications and advanced services …for all consumers at just, reasonable and affordable rates.”

The closest federal policy came to mandating universal access to advanced telecommunications was in 2015 when the FCC placed Internet protocol telecommunications under Title II of the Communications Act of 1934, classifying it as a common carrier utility requiring reasonable requests for service be honored and barring neighborhood redlining. The FCC declined to enforce its regulation adopting the reclassification and ultimately reversed course in 2018, repealing it.

The IIJA did not affirmatively express public policy of universal service. It merely stated findings that access is “essential to full participation in modern life in the United States.” Universal service is described in the legislation by its inverse: a “persistent ‘digital divide’ in the United States.” It charged states receiving planning grants to only determine how long it would take to construct infrastructure providing universal service.

That Americans have not seen universal landline delivered advanced telecommunications reflects a longstanding problem of insufficient political will for policy ensuring fiber would reach most every American doorstep. That would constitute “belt and suspenders” advanced telecommunications infrastructure that would serve well into the 21st century.

Instead, Americans have seen numerous, limited one off subsidies largely directed to investor owned providers with limited capacity to invest. Often that has meant no FTTP belt and only wireless suspenders to reliably hold up the trousers of its connectivity needs over the long term.