Comcast: A Future
How a 40-year defensive perimeter got routed around at every layer simultaneously, and why almost nobody at the top of the cable industry has noticed yet
The cable industry spent twenty years building a state-level pre-emption moat against muni broadband. The replacement is showing up under federal-tribal authority, financed by NTIA grants, strung on poles the cable companies sold to county nonprofits, by crews the cable companies trained and laid off, with chatbots doing the legal research. None of it is reversible. The fiber does not unbuild.
This is a future. This story is fictional, but its coming true, fast.
I. The pole was always the moat
Telephone poles in most of America used to be owned by some combination of incumbent local exchange carriers, electric utilities, and the cable operators themselves. Access to the pole was governed by attachment agreements that were one of the cable industry’s quiet defensive moats. When the incumbent owned the pole, the incumbent could slow-walk attachment requests from competitive carriers, charge attachment fees that made the competitor’s economics worse, require make-ready engineering that took eighteen months to schedule, demand insurance and indemnification terms that smaller carriers could not meet. The pole regime was a soft moat but a real one. It added two to four years of delay to most overbuilders, which was usually enough to make the overbuild uneconomic.
Then the monopoly utility companies sold the poles.
Not all at once. Over 25 years, in waves, on the same financial-engineering logic that produced a dozen other transactions in the same family. The internal pitch was clean: the pole assets were carrying ugly depreciation schedules, unbounded maintenance liability, escalating make-ready exposure under the FCC’s evolving access rules, and regulatory overhead that complicated the rate cases. The CFOs who pushed the divestments got bonuses for it. The boards approved it. The asset class moved.
It moved into county-level not-for-profits. Joint pole authorities. Special districts. 501(c)(12) cooperatives. The structure varied by state. The common pattern was that the new pole owner was a tax-exempt entity governed by a board that was effectively an arm of the local public utility district, the rural electric cooperative, or the county government directly. The pole authority’s mandate was to maintain inventory and manage attachments under nondiscriminatory access rules. The pole authority had no shareholders. The pole authority had no quarterly earnings call. The pole authority’s directors were the same local officials who, in their day jobs, were sponsoring broadband expansion as a top-three civic priority.
The cable industry did not see this as a strategic loss because, from inside the financial-engineering frame, the new structure looked equivalent to the old one. Attachment rates were comparable. The technical processes were similar. What the industry missed was that it had transferred the substrate of the network from an asset class it partly controlled to an asset class permanently aligned with its adversaries. When the tribal carriers and the cooperatives and the PUD carriers came to the pole authority asking for attachments, the pole authority did not slow them down. The pole authority expedited them, because the pole authority’s board members were the people sponsoring the broadband initiatives in their day jobs.
The “open-pole regime” that overbuilders now enjoy across most of rural and exurban America is not the result of regulatory reform. The regulatory reform (FCC one-touch make-ready in 2018, parallel state PUC rules in roughly half the states, federal grant conditions requiring pole-access reform, FCC jurisdiction over tribal disputes) is the visible part. The invisible part is that the pole inventory itself, the physical substrate, is now in friendly hands. The cable industry handed it over for cash. They thought they were getting paid to take a problem off the books. They were getting paid to surrender the high ground.
II. The crews are the same crews
The men in the bucket trucks stringing the fiber that will replace Comcast in rural America are men Comcast trained.
The internal name for the program was “contractor conversion.” Over fifteen years, in waves, the cable industry moved its field workforce off payroll and into contractor relationships. The pitch deck looked the same each time. The company was carrying a workforce with pension liabilities creating accounting drag, workers’ comp exposure that was variable across geographies, a wage structure that did not flex with regional labor markets, and a headcount-to-revenue ratio that the analysts had decided was load-bearing on the valuation multiple.
Each wave moved a chunk of field workforce off the payroll into a contractor relationship. The contractors were small, locally owned shops. Two-truck shops. Five-truck shops. The crew lead was someone who had worked at the cable company for fifteen or twenty years, knew the plant, knew the customers, knew the geography, had been laid off in one of the conversion waves, had taken severance, had used the severance to buy a truck and a bucket lift and an inventory of connectors, and had come back as a vendor. He was the cable company’s only customer in year one. They negotiated rates with him in year two. By year five he had three trucks and four employees, all of them former cable employees, all of them trained on the plant, all of them holding certifications the cable company had paid to certify them for during their employment.
The certifications stayed with the worker. The training investment stayed with the worker. The institutional knowledge stayed with the worker. The cable company had paid for all of it. They had then handed it to a small business owner who was now charging them by the hour to use it.
At the time, the math worked. The contractor’s hourly rate was lower than the cable company’s fully-burdened cost for the same work, because the contractor was not carrying pension obligations, healthcare cost structure, union work rules, or administrative overhead. The savings were real and showed up in operating margin and the analysts liked it.
What the cable industry did not understand at the time, because the framing was financial-engineering rather than competitive-strategy, was that it had taken its specialized labor force and converted it into a general-purpose regional labor market for fiber installation, splicing, and field operations. The contractor was not, in any meaningful sense, the cable company’s contractor. The contractor was a small business in a regional labor market. The contractor would work for whoever was paying.
For ten years, there was no competing buyer. The cable company was the only entity buying field services at scale in rural and exurban America.
That changed when the federal grant money started hitting the rural buildout. The tribal carriers, the cooperatives, the PUDs, the regional fiber overbuilders all needed exactly the same labor that had been trained and then divested. They needed splicers. They needed aerial crews. They needed underground crews. They needed certified fiber technicians who knew the local geography, the local pole inventory, the local right-of-way conventions, and the local permitting offices.
The labor was already there. The cable industry had built it. The new entrants did not have to recruit. They did not have to train. They did not have to certify. They placed an order with the same small contractor at a higher hourly rate than the cable company was paying, and the contractor said yes, because the contractor’s incentive structure was to take the higher-paying work.
The first wave of the rural buildout was effectively staffed by former Comcast employees, working for small contractors Comcast created, paid by federal grant money flowing through tribal and cooperative carriers, to install the fiber networks that were taking Comcast’s subscribers. Every component of the staffing model was something the cable industry had personally constructed. The trucks were the trucks former employees had bought with the severance the cable company had paid them to leave.
The contractor in central Pennsylvania, the former line foreman, has three kids in college now. The cooperative pays forty percent more than Comcast. The math is the math.
III. The asymmetry of expertise has collapsed
The cable industry’s defensive posture against muni broadband was predicated on an asymmetry of expertise.
County commissioners in places with under fifty thousand people are part-time elected officials with day jobs. They are insurance agents and feed store owners and retired schoolteachers. They do not have general counsel. They do not have telecom consultants on retainer. They do not have research staff. When the cable industry’s lawyers showed up with a forty-page memo explaining why the proposed muni broadband project violated the state’s anti-overbuild statute and would expose the county to ruinous litigation, the county commissioner could not evaluate the claim. The commissioner’s options were to pay an outside law firm forty thousand dollars to write a counter-memo (which the county did not have in the budget) or to back down. The commissioner backed down. This was the whole game. The asymmetry of expertise was the moat.
The moat is gone.
A county commissioner in Hazzard County, population seven thousand, can now sit at her kitchen table at nine on a Tuesday night and ask Claude whether there is a federally recognized tribe within a hundred kilometers of her county seat whose trust lands or service area could plausibly anchor a tribal-county broadband partnership under the NTIA Tribal Broadband Connectivity Program, and get back a list with the Colville Confederated Tribes at the top, with a summary of their existing telecommunications authority, with a description of the relevant federal funding streams, with the legal theory under which the partnership would be structured, with the names of the three or four other counties that have done something similar, and with a draft letter to the tribal chair opening the conversation. In ninety seconds. For free. At her kitchen table. With nobody watching.
The commissioner does not need to be sophisticated. She does not need to have read Cohen’s Handbook of Federal Indian Law. She does not need to know what McGirt was or why the Castro-Huerta carve-back matters or how the 2.5 GHz tribal priority window from 2020 worked. She needs to be able to ask a question in plain English and recognize a useful answer when she sees one. That is now the entire skill stack required to route around twenty years of cable-industry state-level lobbying investment.
The diffusion is happening below the surface of anything any cable company’s Government Affairs team monitors. Government Affairs monitors statehouses, lobbyists, ALEC working groups, industry trade press, and FCC dockets. Government Affairs does not monitor the cumulative private chat histories of seven hundred county commissioners in flyover country. Nobody does. The data does not exist in any aggregable form. The aggregate effect is real and is moving fast and is invisible.
If county commissioners have this, so do tribal councils. So do the rural electric cooperatives. So do the small ISPs that the cable industry was historically able to sue into the ground because the small ISP could not afford to litigate. The small ISP can now afford to litigate, because the small ISP’s lawyer has Claude open in another tab and is operating at three times their previous billable efficiency. The small ISP is suddenly viable in a court fight that the cable industry’s lawyers had been pricing as unwinnable for the small ISP. The whole architecture of legal-cost-as-moat is degrading simultaneously across every adversary class the cable industry faces in rural markets.
There is no policy lever the industry can pull. There is no statehouse to lobby. There is no ALEC model legislation that prevents county commissioners from using a chatbot at their kitchen tables. The capability is in the water. The capability is going to get more capable, not less, on roughly a six-month doubling cadence. The asymmetry is gone and is not coming back.
IV. The legal architecture moved while the lobbyists were watching the wrong door
The tribal sovereignty piece is the part that took the cable industry by surprise, even though it should not have.
For thirty years after California v. Cabazon Band (1987) and IGRA (1988), federally recognized tribes ran a specific kind of arbitrage: gaming on trust land that the surrounding states could not run. Casinos required state compacts (which kept the state as a permanent counterparty), required proximity to non-tribal population centers (which most reservations did not have), and saturated their catchment areas. The Seminoles made it work at scale through Hard Rock International. The Pequots made it work for a while and then did not. Most tribes got something between modest and disappointing. Almost none of it was sovereignty-enhancing in the long run, because the compact structure made the state a permanent counterparty.
Telecommunications is structurally different in every dimension that matters.
It does not require proximity to non-tribal population centers. The opposite. The further from population centers, the less competition, because the unit economics of cable and fiber buildout get worse with distance, which is why Comcast and Charter abandoned rural America in the first place. Reservations are sited in places where incumbents do not want to operate. That used to be a curse. In the broadband era it is a moat. Tribes locked out of casino economics by geography are uniquely positioned in broadband economics by the same geography.
It does not require state compacts. The FCC regulates spectrum at the federal level. Tribal telecommunications authority is recognized federally. The state does not get a vote. The state’s twenty-year investment in pre-emption legislation, the entire ALEC apparatus, the careful cultivation of statehouse relationships by the cable lobby, all of it is irrelevant to a tribally-owned fiber network operating under a tribal utility authority on trust land with federal funding. The state cannot tax it, cannot regulate it into uneconomic conditions, cannot license-and-permit-delay it to death. The state is simply not in the chain of authority. The cable industry’s defensive perimeter does not face this direction. The perimeter was built to defend against state-authorized municipal entrants. The actual entrants are federally chartered sovereigns operating with federal money.
It does not saturate. Every household needs broadband. Every business needs broadband. Every tractor in the next-generation precision agriculture stack needs broadband. Every remote worker needs broadband. A tribal fiber network that serves the reservation, the surrounding county, and regional industrial customers (data centers, irrigation systems, oil and gas SCADA, utility telemetry, school districts, county government as anchor tenant) has multiple revenue streams that grow rather than saturate. The middle-mile and backbone capacity, once built, can be wholesaled to other carriers, including the very incumbents who tried to keep the tribe out of the market. Comcast ends up paying the tribe for transit. The reversal is complete and permanent.
It generates jobs that build human capital rather than degrade it. A casino employs dealers, security, hospitality staff, and middle management. None of that transfers outside gaming. A telecommunications utility employs network engineers, fiber technicians, NOC operators, billing systems administrators, and field service crews. That stack is portable, scales with the network, and produces tribal members who can train the next cohort and be hired back into tribal government for water, power, transportation. Casino revenue funded scholarships. Telecom revenue funds an engineering corps. The difference compounds.
It generates federal political capital instead of consuming it. Casino expansion required tribes to fight a defensive political war for thirty years against state governments and anti-gambling moral entrepreneurs and competing tribes. Broadband expansion has tribes operating in alignment with every federal priority of the last three administrations: rural broadband access, digital equity, telehealth, distance education, agricultural modernization, supply chain resilience. The Senate Committee on Indian Affairs likes broadband tribes. The Department of Commerce likes broadband tribes. The USDA Rural Utilities Service likes broadband tribes. The Bureau of Indian Affairs likes broadband tribes. State governments dislike municipal broadband intensely but find it politically impossible to publicly oppose tribal broadband, because the optics of a state government opposing a tribal nation’s exercise of telecommunications sovereignty in service of rural digital equity are unsurvivable.
The federal money is real. The NTIA Tribal Broadband Connectivity Program has put roughly $3 billion into the ground. The BEAD program has tens of billions allocated. The USDA ReConnect program is contributing. The Treasury Capital Projects Fund is contributing. The 2.5 GHz tribal priority window in 2020 gave tribes spectrum that some flipped into network capital and some used to build coverage that radiates well off-reservation. Approximately forty-seven billion dollars of federal infrastructure money got deployed into rural markets over the last several years while the cable industry was defending state pre-emption laws against threats that no longer mattered because the threat had moved.
The internal sentence the tribal leadership should soon be saying out loud, in council meetings and in the closed sessions of the National Congress of American Indians and in the working groups of the National Tribal Telecommunications Association: casinos were a permitted exception to a sovereign disability; telecom is a sovereign capability with no asterisk.
That is the reframe. The casino was the federal government saying we will allow you to do this thing in this narrow way under these conditions because it is in our interest to allow it. The telecom buildout is the tribe saying we have authority that predates the United States, and we are exercising it in a domain where the United States acknowledges our authority, and we are doing it on our timeline with our partners on our terms. The fiber in the ground cannot be revoked. The trust land cannot be revoked. The federal funding flows are statutory, not discretionary. The capability, once built, is permanent.
The tribal carriers are signing wholesale agreements with regional carriers. The federal middle-mile money is funding cross-tribal transit. Within ten years there will be a continuous tribal-anchored fiber network across the rural West, the upper Midwest, and parts of the Southeast. The cable industry’s analysts are missing this because their dashboards do not have a row for backbone capacity controlled by sovereign entities outside our regulatory model. The dashboards reflect what the industry has historically chosen to measure, which is what the industry historically had to worry about. The dashboards are not being updated.
V. The voting block of normies
There is a parallel non-tribal vector running on the same incentive gradient.
The cable industry’s Government Affairs teams call it “the off-cycle problem” internally. For twenty years, muni broadband initiatives were defeated at the ballot box through a reliable playbook: flood the zone with mailers, fund a local astroturf group with a name like Citizens for Responsible Broadband, get the incumbent state legislators (whose campaign contributions had been received) to pre-empt municipal authority at the state level, run the clock. Most states have laws on the books, written by ALEC or its functional equivalents, that ban or severely restrict municipal broadband.
That equilibrium is breaking. Off-cycle muni broadband referenda that used to fail 60-40 against well-funded opposition are now passing 70-30 with the opposition outspending the proponents 50-to-1. It is happening in places where the demographic that turns out for a March Tuesday school board election has decided that giving the incumbent another chance is a worse use of their tax dollars than building their own goddamn fiber. Once the fiber is in the ground, owned by the PUD, run at cost, it does not unbuild. The capital expenditure is permanent. The customer relationship is permanent. The revenue is gone forever.
When a county commissioner cannot win the muni broadband fight on its own, she does not give up. She drives out to the tribal council with a folder under her arm. The county becomes the anchor tenant on a tribally-prime project. Comcast cannot sue the tribe, because suing a sovereign over its sovereign exercise of telecommunications authority is a losing posture politically, legally, and in the press. The Bureau of Indian Affairs is not Comcast’s friend. The DOJ Indian Resources Section is not Comcast’s friend. The Senate Committee on Indian Affairs is very much not Comcast’s friend. The fight is unwinnable for the cable industry on that ground, which is why the cable industry’s lawyers stop sending letters the moment the tribal entity is named on the project paperwork.
For four hundred years, the dominant fact of tribal-county relations in rural America was that the county did not need the tribe and the tribe needed the county for access to non-tribal economic networks. That has reversed. The county now needs the tribe to deliver broadband the county cannot legally build for itself. The tribal council is absorbing the reversal and recalibrating around it.
VI. The financing structure (or, why the connection fee does not matter)
There is a piece of this that the cable industry’s strategic planners have systematically failed to model.
Many of the new fiber networks are not legally utilities. This is a subtle point. A legal utility, in most states, has the authority to issue tax-exempt revenue bonds backed by future ratepayer revenue, which is how municipal water and electric and regional sewage have been financed for a hundred years. Bond proceeds pay for the buildout. Subscriber fees service the bonds. The household pays nothing up front and pays a monthly utility bill calibrated to amortize the construction debt over twenty or thirty years. That is the historical model.
The new fiber networks largely cannot do that, because in most states they are not classified as utilities under the relevant statutes. They are classified as cooperatives, as special purpose entities, as public-private partnerships, as broadband authorities. The utility classification was historically a regulated monopoly classification, and the new networks did not want it because it would have brought rate-of-return regulation and the rate case overhead that comes with it. They wanted the operational flexibility of being non-utility entities. The trade-off was that they could not finance the buildout through traditional muni revenue bonds.
What they did instead was a hybrid. The bulk of construction capital came from federal grants (NTIA TBCP, BEAD, USDA ReConnect, Treasury Capital Projects Fund, and a long tail of smaller programs), which covered the middle mile, the head-end, and a substantial fraction of the last mile in low-density areas where the unit economics did not pencil out otherwise. The remainder of the last-mile cost (the actual drop from the pole to the home) is recovered through a connection fee paid by the subscriber at signup. The connection fee is typically four hundred to fifteen hundred dollars depending on geography and carrier, and it covers the marginal cost of the install plus a contribution to operations reserves.
The connection fee looks, on paper, like a barrier to adoption. It is a real upfront cost in a market where the incumbents charge nothing upfront and recover the install cost through twelve months of monthly billing. If the question is whether an existing homeowner will pay an eight-hundred-dollar fee to switch from Comcast to the fiber co-op, the answer in many cases is no, or not immediately. The friction is real. The cable industry has been quietly counting on this friction as a defensive feature.
The friction does not exist in new construction.
In every market where a fiber network has been built and is operating, the home builders and the developers have noticed that fiber is a competitive amenity in selling new homes, and they have started pricing the connection into the build. The fiber drop is being installed alongside the water tap and the sewer connection and the electric service, by the same crews working from the same construction schedule, with the same coordination meetings between the developer and the local utility authorities. The connection fee is being paid by the developer as part of lot improvement cost, the same way the developer pays for the curb cut and the storm drain tie-in. The fee gets capitalized into the lot price, capitalized into the home price, amortized into the mortgage over thirty years.
The home buyer never sees the connection fee as a discrete decision. The home buyer sees a house that costs four hundred thousand dollars instead of three hundred and ninety-nine thousand two hundred. The eight hundred dollars disappears into the price stack along with the upgraded countertops. The home buyer moves in and the fiber works. The home buyer does not call the cable company, because the home buyer never had a relationship with the cable company. The home buyer’s address was, from the moment the developer pulled the permit, off the cable company’s service map and onto the new entrant’s service map.
The cable industry does not even get the chance to lose this customer. The customer was lost at the planning department of the county building, eighteen months before the house was framed.
The new construction rate in suburban and exurban tiers is the demographic engine of net household formation. The cable industry has been counting on that engine to backfill its subscriber base as existing households age out. The backfill is not happening anymore. The exurban builders have been pricing fiber into new construction for at least three years. The suburban builders are starting now in markets where fiber has reached the suburban edge. The builders are ahead of the buyers. The builders are pulling the suburban market into the new regime in advance of the suburban buyers becoming aware that the regime exists.
VII. The machine
There is a piece of this that almost nobody outside the field operations community has been tracking.
A manufacturer has been refining a single-operator micro-trenching machine for about ten years. The published specification is eight hundred meters per shift in standard suburban asphalt. For comparison: traditional underground fiber installation in suburban environments runs at fifty to one hundred meters per shift with a four-to-six person crew, depending on geology, permitting, and existing utility congestion.
This can deliver ten times the throughput at one-fifth the labor cost. A fifty-fold improvement in labor productivity for the underground portion of the buildout.
The machine knifes down thirty-five centimeters, which is below the frost line in most of the country. It places a twenty-five millimeter conduit with mutlplechannels inside. The slot self-seals through asphalt displacement and a thin sealant bead laid by the trailing edge of the machine. The total disturbance to the road surface is a thin orange line that weathers flush within a few weeks of vehicle traffic. One operator. One pass.
The conduit is approximately $2.40 per meter at current production volumes, roughly forty percent less than equivalent microduct at similar channel counts. The price is falling as production scales. The conduit is now cheaper than the labor cost it displaces by a substantial margin. Conduit cost is no longer the binding constraint. Labor productivity is no longer the binding constraint. The binding constraint, in suburban environments where such installs can operate, is permitting and right-of-way access, and the permitting environment for micro-trenching is substantially easier than for traditional underground installation, because the surface disturbance is minimal, the depth is below the frost line, the system does not interfere with existing buried utilities at its operating depth, and the restoration is complete on the same day as the cut.
The conduit is placed. The fiber is placed. Conduit is sunk capital but conduit is cheap.
VIII. The federal anchor
Federal procurement is the largest single revenue stream the commercial telecommunications industry has historically depended on. GSA’s Enterprise Infrastructure Solutions framework, the Defense Information Systems Agency’s parallel track for DoD, and the intelligence community’s separate arrangements consume substantial commercial telecommunications capacity in the aggregate. The contracts have been treated as durable. They are not durable.
The National Tribal Telecommunications Association can offer the federal government a multipoint peering agreement that the commercial carriers cannot match. The reasons it cannot be matched are not, fundamentally, about price.
The Association’s member are sovereign entities under federal recognition. Their ownership is not subject to foreign acquisition. Their operational control is not subject to foreign influence. Their workforces are domestic by definition. The federal government has been quietly worrying about foreign-ownership exposure in the commercial carrier base for fifteen years. The worry has accelerated as the geopolitical environment has hardened. The commercial carriers do not have a clean answer to the worry, because their capital structures are global, their supply chains are global, their workforces are global. The Association has a clean answer.
The Association networks were built new, with federal grant capital, under federal supply-chain rules. The hardware stacks are largely or entirely compliant with the Trusted Communications and BABA requirements that have been getting increasingly aggressive. The Association can certify supply-chain compliance in a way that the commercial carriers cannot, because the commercial carriers have an installed-base retrofit problem and the Association does not.
A federal procurement officer who accepts the Association proposal can frame the acceptance as advancing tribal economic development, increasing competition in the carrier market, supporting rural broadband infrastructure, and improving value-for-money in federal procurement. A federal procurement officer who rejects the proposal in favor of continuing with the commercial carriers will be asked, by a congressional staffer at some point in the next budget cycle, to explain the rejection on the record. The explanation will be uncomfortable. The procurement officer will be defending a status quo without a defensible political constituency. The political path of least resistance is acceptance, and the federal procurement apparatus walks the political path of least resistance regardless of what the formal procurement criteria say on paper.
By aggregating federal telecommunications spending across the Association member carriers, the peering agreement creates a revenue stream that funds the next phase of the Alliance’s buildout: metropolitan fiber rings anchored on federal-customer facilities. Federal facilities are everywhere. Military bases, federal courthouses, federal office buildings, VA hospitals, federal research facilities, post offices, IRS regional centers, Social Security offices. They are in every metropolitan area. They are typically located in the urban and suburban core. They are anchor tenants whose anchor commitment justifies the construction of metropolitan fiber rings that pass through the urban and suburban core on the way to them.
The peering agreement, by giving the Association the federal anchor in every metropolitan area, gives the Association the operational and financial basis for building urban and suburban fiber capacity in markets where the rural-to-urban diffusion would otherwise have been slow. It collapses the timeline on the urban migration. The cable industry’s strategic planners had been forecasting the urban transition at ten to fifteen years. With the federal anchor, it is five to seven, possibly faster.
IX. The IPv6 multicast gap nobody at the top of the cable industry has heard about
When the FAST operators (Pluto, Tubi, Roku Channel, Xumo, Freevee, Local Now) sits down with the Tribal Association and a coalition of second-tier carriers to design a controlled-access multicast network, they do not invite Comcast.
There are political and commercial reasons for the exclusion, all of them sufficient on their own. There is also a technical reason, which is the one that does not get discussed in public. The technical reason is that Comcast cannot do it.
The cable industry has IPv6 deployed at the access layer in a fraction of its footprint, with native dual-stack support in roughly that fraction, and carrier-grade NAT bridging the IPv4 islands in the remainder. The IPv6 deployment is for unicast traffic. Multicast capability at the IPv6 layer is essentially nonexistent inside the cable industry. There is IGMP multicast capability at the IPv4 layer, built fifteen years ago to support the legacy video distribution architecture, that has not been meaningfully upgraded since. The legacy IPv4 multicast architecture is not interoperable with the IPv6 multicast architecture that any modern multi-party multicast system is built on. The legacy architecture is also operating on equipment that is at end-of-life across substantial portions of the network.
The capabilities a modern carrier-scale IPv6 multicast deployment requires (PIM-SSM source-specific multicast routing, MLDv2 listener discovery, segment routing for traffic engineering, operational tooling to manage all of it at carrier scale) do not exist at production-grade quality in any meaningful portion of the cable industry’s footprint. There are lab implementations in engineering test environments, sometimes built on individual engineers’ initiative against budgets quietly hidden in operations expense lines. They have not been productized. Productization would require eighteen to thirty months of capital deployment and operational integration before the cable industry could offer carrier-grade IPv6 multicast service to a customer.
The peering and interconnection architecture is optimized for unicast east-west traffic with the major content providers and is not configured for multi-party multicast distribution. Reconfiguring it would require renegotiating peering agreements stable for a decade. The DOCSIS 4.0 deployment is behind the participating second-tier carriers’ fiber-deep deployments by a substantial margin in markets where the deployments overlap, which means the cable industry’s last-mile capability cannot match the multicast performance characteristics that a modern multicast architecture requires. The network operations tooling does not have multicast monitoring and troubleshooting capabilities at the depth required. The network operations workforce has not been trained on multicast operations at the depth required.
The cumulative effect is that the cable industry could not have been a credible carrier-tier participant in any modern multicast consortium even if the political and commercial considerations had not excluded it. The exclusion was overdetermined. The political and commercial reasons were sufficient. The technical reasons were also sufficient. Either category would have produced the exclusion on its own.
The consortium’s technical leadership knew this because the consortium’s technical leadership has been participating in the same standards working groups (IETF, MEF, CableLabs, IEEE 802) and reading the same trade press and observing the same public peering measurements and BGP table data and academic research papers that have measured large carrier networks for the last decade. The cable industry is not a black box to the consortium. The consortium has known what the cable industry’s capabilities are, in technical detail, for years. The consortium did not have to investigate. The consortium had to look at what was already public.
What was already public was that the cable industry had not invested in IPv6 multicast capability at carrier scale, had not invested in the operational tooling that would support such capability, had not invested in the workforce training that would operate such capability. The consortium structured itself around the carriers that had been making those investments. Those carriers were the second-tier carriers and the tribal carriers who had been building greenfield fiber networks under the federal funding programs and who had built those networks on modern architectures that included IPv6 multicast as a foundational capability rather than as an afterthought.
The consortium did not have to find the right carriers. The consortium identified the carriers who were technically ready, and those carriers were the ones who had built the new networks, and those carriers were not Comcast.
X. The vertical stack closes
The carrier side is only half of the multicast play. The other half is the consumer electronics ecosystem, and the consumer electronics ecosystem has already committed.
Amazon is providing the consumer-side device platform through Fire TV and the operating system layer through Vega OS, which has been replacing Fire OS on new Amazon devices and is now licensed to a growing number of third-party television manufacturers. Vega OS includes native support for the multicast streaming protocols that any modern controlled-access multicast architecture is built on. Amazon has been quietly developing this support for at least three years. The native support means Vega OS devices on a multicast-served network receive multicast streams without user-side configuration, with channel switching latency in the range of two to four hundred milliseconds, comparable to legacy cable television and substantially faster than current unicast streaming. Amazon Ads provides the unified ad inventory management layer with preferential terms for participating FAST operators.
Apple is providing native support for the multicast protocols in tvOS and in the AirPlay 2 protocol stack. Apple TV devices and AirPlay-compatible smart televisions on a participating network receive multicast streams natively at the same low latency. Apple is providing App Store distribution for the FAST operators’ applications with preferential placement in the tvOS App Store. Apple is providing privacy-preserving identity infrastructure through Sign in with Apple and Private Relay, which enables household-level subscription management without exposing individual viewer identity to the operators or carriers. Apple does not take a percentage of the consortium revenue; Apple takes the strategic benefit of having Apple devices be the preferred consumption endpoint on the new architecture.
Sony is providing the television manufacturing partnership for the high end. Bravia televisions ship with multicast integration in firmware, with dedicated UI elements for the consortium’s content, and with hardware optimizations on picture and audio quality that generic streaming devices cannot match. Sony Pictures Television is a content provider on preferential terms, which gives the participating channels access to Sony Pictures library content that competing distribution architectures will not have, or will have on less favorable terms. Sony is the first major studio to commit. The other major studios will follow on a twelve-to-thirty-six-month lag.
LG is providing the television manufacturing partnership for the mid and low ends, which is the volume part of the television market. webOS, which ships on LG televisions and is licensed to a growing number of third-party manufacturers including some of the major Chinese brands not subject to federal supply-chain restrictions, has been updated to include native multicast support in the firmware version shipping on all new LG televisions starting with the 2025 model year. The integration includes a dedicated application that aggregates the participating FAST channels into a unified browse-and-watch experience that resembles legacy cable television in its UI metaphors. This is deliberate. The consortium has done extensive consumer research on the user experience preferences of the demographic most likely to adopt the new architecture: the cord-cutter and cord-never demographic that left or never had cable but retains positive associations with the channel-flipping browse experience that legacy cable offered. The webOS application reproduces that experience. The user experience will feel, to the consumer, like a return to cable television but with no monthly bill.
Amazon Fire TV alone has approximately two hundred million active devices in North American households. Apple’s tvOS and AirPlay-compatible installed base is approximately seventy million. Sony’s Bravia installed base is approximately forty million. LG’s webOS installed base is approximately one hundred and twenty million. The four installed bases overlap substantially, but the union covers roughly two hundred and fifty to three hundred million household device endpoints. The integration is being deployed to existing installed bases through firmware updates. Consumers do not need to buy new hardware. The integration arrives on existing devices through ordinary software updates.
Samsung is being deliberately kept out of the initial announcement to build pressure on Samsung to negotiate from a weaker position. The strategy will work. Samsung will join, on terms less favorable than the four founding manufacturers received. The consortium will use the Samsung capitulation as a reference point for the remaining holdouts. By twenty-four months from now, every major consumer electronics manufacturer in the streaming television category will be a participant.
The architecture delivers, to the consumer, a streaming television experience materially better than any non-participating service can deliver. Faster channel switching. Higher quality video. Lower latency. Better synchronized audio. More reliable performance during peak hours. Cleaner integration with existing devices. More predictable advertising. Stronger privacy protections. The advantages are individually small. The cumulative advantage is the kind of qualitative gap consumers can perceive in side-by-side comparisons.
The cable industry is not in this stack. The cable industry cannot get into this stack on the carrier side, for the technical reasons. The cable industry can only enter as a content provider through NBCUniversal’s licensing relationship, on terms that are commercially less favorable than what Sony Pictures negotiated as a founding participant, because by the time the cable industry approaches the consortium, the reference points will already be set, and the cable industry will be approaching from a position of weakness.
XI. What the disposition curve looks like once you back the noise out
Watch the rural disposition data for any major cable operator over the trailing thirty-six months. The “terminated” line (households that have completely disconnected, not cord-cut but retained broadband, not downgraded, gone) has gone from a flat low-single-digit-percent baseline (representing the normal churn from death, relocation, and bankruptcy) to a curve that is no longer flat.
The terminated households are not churning to a competitor in the way the cable industry uses that word. They are not switching to AT&T or to a regional cable operator or to a fixed wireless ISP. If they were, retention tools would work, and retention tools are not working. Pricing levers are not working. Bundle adjustments are not working. Win-back campaigns are not working. The reason none of it works is that the household on the other end of the call is no longer in the cable industry’s market. They are on a fiber connection delivered by a tribal telecommunications authority, or a county PUD, or a rural electric cooperative, or some partnership of the three. The price point is something like $40/month for symmetric gigabit fiber, with no contract, no equipment fee, and no annual increase. The cable industry cannot match the price because the new entrant’s cost of capital is federal grant money and the cable industry’s is equity. The cable industry cannot match the service quality because the new entrant’s plant is new fiber and the cable industry’s is twenty-year-old coax that the operations team has correctly noted has not had the capex case to refresh.
The contagion to the next density tier is not a scenario. It is the trajectory. The same mechanism that took the lowest tier will take the next, on a lag of twenty-four to thirty-six months, because the same federal funding programs that built the lowest-tier backbone are now being deployed into the next-tier markets, the same partnership structures are being replicated, and the tribal carriers that proved the model in the lowest tier are now offering managed services to other tribes starting later. The diffusion is faster than the cable industry’s response cycle. It has been faster for at least five years. The cable industry did not see it because the cable industry was not looking in the places where the diffusion was happening. The places where the diffusion was happening were tribal council meetings, county commissioner meetings, and the private chat sessions of approximately seven hundred elected officials and tribal administrators who, beginning around 2024, started using AI tools to research their options.
The suburb migration is the next phase. The exurbs have been doing the work for three years. The suburban builders are watching the exurban builders. The information has not fully diffused to the suburban consumer yet. That is the only reason the suburbs still look stable on the disposition data. The suburbs do not yet know that twenty miles outside their boundary, in the exurbs, the same broadband service that the suburbs are paying ninety-five dollars a month for is available for forty dollars a month at five times the speed. When the suburbs learn this, the suburbs will respond, and the suburbs will respond fast, because the suburban population is dense, organized through homeowner associations and school district politics and neighborhood social networks, and capable of generating sustained political pressure on municipal governments in ways that rural populations could not.
The suburbs will demand the open-pole regime. The municipalities will fold. The cooperative carriers and regional fiber overbuilders will move in. The economics will not be quite as good as the rural economics, because the suburban entrants will not have the same federal grant pool, but the economics will be good enough, because the cost-of-capital advantage and the operating-cost advantage of new fiber over twenty-year-old coax is large enough to support a forty-to-fifty-dollar price point even without grant subsidy, and suburban density is high enough that the unit economics work without subsidy.
There is no defense. The cable industry’s defenses were all built at the state and federal levels. The suburban fight will be won at the municipal level, where the cable industry has no presence and no relationships and no playbook.
XII. Antitrust evaporates exactly when the company stops being a monopolist
The cable industry’s regulatory framework for the last twenty-five years has been organized around managing antitrust exposure. Sherman Act exposure, Clayton Act exposure, FTC enforcement priorities, the state attorneys general consortium, European regulatory posture for international segments, NBCUniversal consent decree compliance. The exposure has been real. The framework has been correctly built around it for the period it was relevant.
The framework’s premise is dissolving in real time. The premise was that the cable industry held substantial market power in the residential broadband and video distribution markets in its service territories. The premise was true. It became more true through the 2010s as NBCUniversal got acquired and as the broader cable industry consolidated. The exposure was managed through behavioral consent decrees, divestitures, pricing discipline, careful avoidance of conduct that would trigger structural remedies.
The cable industry is not going to be a monopolist in residential broadband in five years, because it is not going to be the dominant carrier in most of the geography that the residential broadband market covers. It is going to be one carrier among several, in retreat from the rural and exurban and suburban geographies, holding a defensible position in the urban core, transitioning to a wholesale infrastructure business that sells transit to the carriers that replaced it. A wholesale infrastructure business selling transit in a market with multiple competing carriers and federally subsidized new entrants is the opposite of a market power position. It is a competitive infrastructure provider in a market with structural oversupply on the demand side, because the new entrants are price-takers on transit and they have multiple sourcing options. Federal infrastructure money has built more middle-mile capacity than the market needs in most regions.
The video distribution side is even more straightforward. The market has been in secular decline for the entire period the antitrust framework cared about. Cord-cutting was the first wave. Complete-disconnects are the second. The video distribution market is not going to exist in its current form in five years. Streaming providers have absorbed the function. NBCUniversal’s content position is real and durable, but the distribution leg of the market-power story is being eroded.
The antitrust exposure profile of the cable industry will look fundamentally different The agencies will take an additional eighteen to twenty-four months to update their internal models, because the agencies update on a lag. There will be a window during which the cable industry is treated as a monopolist by an enforcement apparatus that has not yet noticed it is no longer a monopolist in the markets the apparatus cares about. The cable industry’s legal departments will be running a deregulatory campaign against constraints that no longer have a reason to exist, and the campaign will succeed, because the underlying market position no longer justifies the burden.
The regulatory framework that constrained the cable industry’s behavior in ways that, in retrospect, served the cable industry’s interests, by preventing it from making strategic mistakes that the underlying market conditions would have rewarded in the short run and punished in the long run, is going to be removed at exactly the moment when its discipline is most needed. The bundling restrictions, the must-carry obligations, the network neutrality obligations, the price transparency requirements, the various behavioral commitments accumulated over the years, all of them constrained the industry. They also, by constraining it, prevented it from accelerating the conditions that produced the new entrants. If the industry had been free to bundle aggressively, to discriminate in interconnection, to extract maximum surplus from the rural and exurban markets, it would have done so, because the financial-engineering framing inside the industry would have rewarded the conduct. The conduct would have produced the political backlash that is now producing the federal infrastructure money that is now producing the new entrants. The regulatory framework slowed the political dynamics that would have eventually produced the same result.
As the regulatory framework relaxes, the financial-engineering framing inside the cable industry is going to push for the conduct the framework previously prevented. The bundling will get more aggressive. The interconnection terms will get harder. The price increases in the captive markets will accelerate. The conduct will be locally rational on a quarterly earnings basis, and it will be strategically catastrophic on a five-year basis, because it will reproduce the political conditions that produced the rural buildout, in the urban markets where the cable industry still holds position, and the urban political response will be faster and more sophisticated than the rural political response was, because the urban political response will be informed by the rural precedent and will have AI-mediated organizational capacity.
There is nowhere to retreat after the urban displacement. The wholesale repositioning is the floor.
XIII. What this is, in one sentence
A 21st-century institution dies when every layer of its competitive moat (legal, physical, financial, technological, demographic, regulatory, expert) gets routed around simultaneously by adversaries the institution is structurally incapable of recognizing as adversaries.
Comcast is not stupid. Comcast is not asleep. Comcast is operating exactly as a publicly-traded American cable company in the year 2026 is structured to operate. The conduct that produced the current condition was, at every step, locally rational. The contractor conversion was rational. The pole divestment was rational. The capex deferral on the rural plant was rational. The state-level lobbying investment was rational. The choice not to deploy IPv6 multicast at carrier scale was rational. Each decision was made by someone whose comp was tied to a metric the decision optimized. None of them woke up one morning and decided to lose rural America to a coalition of tribal sovereigns, rural electric cooperatives, county PUDs, and Japanese trenching equipment.
They each, individually, did the locally rational thing. The locally rational things composed into a system that loses to a coalition of tribal sovereigns, rural electric cooperatives, county PUDs, and Japanese trenching equipment.
This is the structure of every 20th-century institutional collapse currently underway. The same shape recurs. The hospital systems are watching the same curve, on a longer timeline, with different specifics. The legacy banks are watching it. The legacy insurers. The law firms. The consulting firms. The universities. Every institution whose moat depended on expertise asymmetry, on legal-cost-as-moat, on physical-substrate ownership, on workforce capture, on regulatory arbitrage with a state government, is on the same curve. Most of them have not noticed. The ones that have noticed are doing what the cable industry did, which is treating it as an operations problem until it is too late to treat it as anything else.
The cable industry is the leading indicator because the cable industry is where the federal funding was deployed first and where the diffusion ran fastest. The same mechanism is going to take the urban consumer markets next. Then commercial. Then enterprise. Then the institutions. The timeline is fifteen to twenty years end to end. Year seven is roughly now.
The fiber goes in the ground.
The fiber does not unbuild.