A proposed framework for Essential Communications Infrastructure: how to fund it, who should pay, and what rural America is entitled to expect from the networks of the next thirty years.
Two articles ago I laid out how the Universal Service Fund got here. The contribution base has collapsed. The factor is projected at 42.3%. The broadband industry that replaced voice pays nothing.
Article 01 The fund that brought communications to rural America is running out of money -
The second article made the case that reforming programs built on a stressed foundation is not the same as fixing it. Washington is spending enormous energy rearranging the deck furniture while the ship is in dangerous waters.
Article 02 Rearranging Deck Chairs -
This article is different. This one proposes something.
I want to be clear about what this is and what it is not. It is an operator's view of what a framework built for the next thirty years should look like. It is not a rulemaking proposal or a legislative draft. I have spent twenty-five years running networks in rural America. I know what it costs to serve communities that no national carrier will touch. I know what the current system does and does not make possible. This is what I would build if I had the chance.
I am calling it Essential Communications Infrastructure. The name matters. Broadband is a technology descriptor. Infrastructure is a policy commitment. What rural America needs is not a better broadband program. It needs a framework that treats communications the way we treat roads and water. Essential. A shared obligation. Something every community deserves regardless of geography.
Start with the Problem
There are two problems here and they are related but not the same.
The first is the contribution base. The voice revenue pool that funds USF has been shrinking for twenty years. The factor required to sustain the fund on what remains just hit a projected 42.3%. That number keeps going up because the base keeps going down. You cannot fix that by restructuring the programs that sit on top of it. You have to fix the base.
The second problem is harder to talk about in policy circles but anyone who has run a rural network understands it immediately.
My network carries a lot of Netflix. And YouTube. And Amazon Prime. And Disney+. My subscribers want those services and I deliver them. To do that I pay for upstream bandwidth capacity sized to handle the peak demand those platforms generate. Netflix alone accounts for roughly 15% of global internet traffic. My capacity costs reflect that.
Netflix charges my subscribers for access to its content. I pay for the capacity to deliver it. Netflix pays nothing toward that infrastructure.
Every link in the delivery chain charges for what it provides. The content creator charges the subscriber. The CDN charges the content creator. The transit provider charges the CDN. The rural operator absorbs a cost driven by all of them and recovers none of it from the entities creating it.
USF was designed to close that kind of gap. A framework that addresses only the contribution base without addressing cost causation solves half the problem. I want to solve the whole thing.
The Governing Principle
The 1934 Act set a goal: make communications service available to all Americans with adequate facilities at reasonable charges. The 1996 Act reaffirmed it and extended the commitment to advanced telecommunications services. Section 254 required equitable and nondiscriminatory contributions from all providers of interstate telecommunications services.
That language pointed toward something that was never built. Here is what building it looks like.
Every entity that generates revenue from U.S. internet connectivity shares in the obligation to fund the infrastructure that serves everyone. It does not matter whether you call your service broadband, content delivery, cloud computing, or streaming. If you make money from the fact that rural Americans are connected, you contribute to the cost of connecting them.
The Contribution Framework
Two tiers. Simple in principle. Administrable in practice.
Tier One: Broadband Access Revenue
Every provider selling broadband access to U.S. subscribers contributes a percentage of that revenue to the fund. ISPs. Wireless carriers. Cable operators. Satellite providers. Everyone selling access.
The base is approximately $300 billion annually. Fixed broadband revenues exceed $104 billion. Wireless data revenues approach $200 billion. At 2 to 3%, Tier One generates $6 to $9 billion. That rate is a fraction of the 42.3% currently extracted from voice revenues. As broadband revenue grows, which it will, the rate required to sustain the fund goes down. That is the opposite of what the current system does.
This tier requires Congressional action to classify broadband access revenues as assessable for universal service purposes. That is the threshold step. It is also the most straightforward argument to make: Section 254 required equitable contributions from all providers. The providers making money from broadband access should be among them.
Tier Two: Upstream Capacity Purchases
This is the tier that addresses cost causation directly.
Every entity above a defined threshold that purchases upstream internet capacity at U.S. interconnection points contributes based on that purchased capacity. Transit circuits. Internet exchange point ports. Dedicated interconnection agreements. Any contracted arrangement that enables a content provider, CDN operator, or large-scale distributor to deliver traffic to U.S. residential networks.
The contribution is on the contracted capacity, not actual traffic. A 100 Gigabit Ethernet transit port at a major U.S. internet exchange currently costs approximately $0.05 per Mbps per month. The USF contribution on that port would be a small fraction of that cost. Not material to a company the size of Netflix or Google. Meaningful to the fund.
Netflix. YouTube. Amazon. Meta. Apple. Microsoft. These are not hypothetical contributors. They are the identifiable drivers of the capacity costs rural operators bear every day. A de minimis threshold set at 10 Gigabits of U.S.-facing contracted capacity would capture the major platforms while leaving small businesses and individual developers outside the obligation entirely.
I have heard the counterargument. Would this drive content providers to route traffic offshore to avoid the assessment? The practical answer is no. These companies have built their competitive advantage on proximity to U.S. subscribers. Latency is not negotiable for commercial content delivery. Restructuring global CDN infrastructure to avoid a contribution that represents a fraction of a percent of revenue is not a rational business decision. The cost of avoidance exceeds the cost of compliance by orders of magnitude.
Netflix charges my subscribers for access to content that my network delivers. My network pays for the capacity to deliver it. Netflix pays nothing toward that infrastructure. This tier closes that loop.
Broadband access revenue estimates based on fixed broadband revenue of $104.8B (ResearchAndMarkets, November 2025) and wireless data revenue estimates. Tier Two base not yet fully quantified pending FCC capacity survey. Total target reflects current USF disbursement level of $8.5B annually.
Two Funds, Not One
The current high-cost program tries to do two different things with one mechanism. Build infrastructure. Sustain operations. Those are not the same objective and they should not be funded the same way.
Fund A builds things. Fiber-first. Thirty-year useful life. Copper networks get support to upgrade to fiber, not to stay copper. Fixed wireless and satellite do not qualify here. They are complements to terrestrial infrastructure, not substitutes for it. Fund A answers one question: what does it cost to build a network that serves this community for the next generation?
Fund B sustains things. It covers the ongoing gap between what subscribers in high-cost territory can reasonably pay and what it actually costs to maintain and operate the network. A March 2025 survey of nearly 270 NTCA members found rural operators receive an average of more than $70 per month per broadband subscriber in USF support just to cover capital recovery, loan repayment, and operating expenses. That is not excess. That is what it costs to serve rural America on an ongoing basis. Fund B keeps that math working.
Congress sets the total. The contribution rate floats to meet it. Not quarterly. On a five-year cycle tied to a broadband revenue and capacity survey. No more quarterly factor volatility. No more budget control proceedings that create annual uncertainty for operators trying to plan capital deployments. The fund is stable because the base is stable and growing.
Who Gets Support and Who Does Not
To be qualified for support under this framework requires physical infrastructure in the community. Not coverage. Presence. We can call it an Essential Communications Infrastructure Provider - ECIP - the creative person out their can turn this into EPIC
A satellite provider can cover every address in rural America from orbit. That is coverage. It is not presence. It does not mean the provider has built anything in the community, employs anyone there, or is accountable to the people it serves in any meaningful way. A satellite provider can raise prices, reduce service, or exit the market without a build obligation, a service requirement, or a community to answer to.
The regulated terrestrial operator who loses support because a satellite provider entered the market cannot exit. Cannot raise prices without consequence. Cannot reduce service without violating its obligations. That asymmetry has to be addressed in how eligibility is defined.
Satellite providers contribute to the fund under Tier Two. They do not receive Fund B support as a substitute for terrestrial infrastructure. In locations where fiber deployment costs exceed any reasonable support level, satellite and fixed wireless serve a real role. A provider willing to serve those locations can opt in to maximum support in exchange for accepting a full service obligation. That is a business decision, not a default outcome.
The 90% Line
I want to be honest about something that most policy frameworks are not honest about.
Not every location can be served to the same standard. The cost curve has a long tail. A framework that pretends otherwise is not a policy. It is a promise that cannot be kept.
The framework sets a cost threshold covering 90% of locations. At or below that line, the build obligation is clear. Fiber. Supported by Fund A. Sustained by Fund B. The economics work.
Above that line, the 10% of locations where deployment costs exceed any standard support level, the framework sets maximum support and requires an affirmative opt-in. A provider evaluates the support and the obligation and decides. Build fiber. Deploy fixed wireless. Partner with satellite. The technology choice belongs to the provider. The obligation to serve belongs to whoever raises their hand. No one is forced. No location is abandoned without someone having the opportunity to serve it.
That is honest. It is also more durable than a framework that overpromises and underdelivers.
Three Years to Get There
The transition happens over three years. That timeline is not arbitrary. Operators are currently in the middle of BEAD and Enhanced ACAM build programs with real capital commitments and real timelines. They need certainty. A longer transition creates more uncertainty, not less.
Year one: new contribution obligations phase in at one third of the full rate. Existing support continues. Operators begin planning for the new framework.
Year two: two thirds. Fund A and Fund B begin operating alongside existing programs. Existing recipients get transition support.
Year three: full framework in effect. Voice-based assessment sunsets. Existing recipients move to Fund B under the new eligibility criteria. Transition support bridges any gap.
The 2011 ICC Transformation Order used a similar phased approach. The industry knows how to execute a structured transition. What it cannot absorb is an open-ended one.
What This Does Not Solve
This framework requires Congressional action. The FCC cannot reclassify broadband access revenues as assessable through rulemaking alone. Tier Two needs either legislation or a legal theory that withstands scrutiny. The proposal here is the target. Legislation is the path.
It does not resolve every implementation detail. Contribution rates, cost model methodology, de minimis thresholds, transition support levels all require FCC rulemaking after Congress acts. I am not pretending this article substitutes for that process.
And it does not guarantee fiber to every door on any particular timeline. BEAD and Enhanced ACAM are already doing that work. This framework sustains and supplements it. It does not replace it.
Why This and Not What Washington Is Doing
The four proceedings I documented in the second article are each trying to optimize programs sitting on a contribution mechanism that cannot sustain them. That is the wrong sequence. You do not restructure the programs before you fix the foundation.
I have learned this lesson running networks. When the underlying infrastructure has a problem, you fix the infrastructure. You do not redesign the services running on top of it and hope the problem goes away.
The Essential Communications Infrastructure framework fixes the foundation first. A stable and growing contribution base. Two funds with clear purposes. Eligibility tied to presence and obligation, not just coverage. Honest acknowledgment of where the limits are and a transparent mechanism for the hardest locations.
The 1934 Act built something for its era. The 1996 Act built something for its era. Section 254 pointed toward what the next era required. That work was never finished.
Finishing it is not a radical idea. It is the continuation of a ninety-year commitment.
Is your organization prepared for what happens if that commitment is not honored? And what would you build if you had the chance?
Sources cited in this article
Communications Act of 1934, 47 U.S.C. Section 1, fcc.gov
Telecommunications Act of 1996, Section 254, 47 U.S.C. 254, congress.gov
ResearchAndMarkets.com, "United States Telecom Operators Country Intelligence Report," November 2025 (fixed broadband revenue figure of $104.8B)
TeleGeography, IP Networks Research Service Executive Summary, 2025 (100 GigE pricing at $0.05 per Mbps per month)
Sandvine / industry estimates: Netflix approximately 15% of global internet traffic
NTCA, Comments in WC Docket Nos. 25-311 and 25-208, May 2026 (March 2025 NTCA member survey, $70/month per subscriber USF support figure)
Broadband Breakfast, "Universal Service Fund Contribution Factor Projected to Reach Record 42.3%," June 2026
FCC, High-Cost Program Reform NPRM, WC Docket No. 26-96, April 2026
FCC, 2011 USF/ICC Transformation Order, WC Docket No. 10-90 (transition timeline precedent)