In 1877, Thomas Doolittle strung the first hard-drawn copper telephone wire in Ansonia, Connecticut, replacing the iron lines that had carried Alexander Graham Bell’s earliest calls. Nearly 150 years later, most of the country is finally retiring the last copper in its telephone networks, replacing it with fiber-optic cable and wireless connectivity.
California, naturally, has other ideas.
Much of the state’s copper network remains in place—powered, maintained, and protected by rules written for a monopoly telephone era that no longer exists. AT&T still provides old-fashioned “plain old telephone service” (POTS) to roughly 3% of households in its California territory. Yet the company spends about $1 billion a year keeping that network alive.
The reason is not consumer demand. It is state regulation. And the costs do not stop at California’s border. They fall on customers across the country.
The Federal Communications Commission (FCC) now has a chance to fix that. AT&T has asked the FCC to declare that, once the agency authorizes POTS discontinuance under federal law, California’s carrier-of-last-resort (COLR) requirements are preempted.
The Hotel California of Telecom
California’s COLR regime dates to the mid-1990s, when the state opened local phone markets to competition but worried some customers might get stranded in the transition. It designated incumbent carriers as “carriers of last resort,” obligating them to provide “basic service” to every residential household, and to serve every business on request, throughout their territories.
The state calls these rules “technology neutral.” In practice, they force incumbents like AT&T to keep legacy copper networks running. The required elements of “basic service”—directory assistance, white-pages listings, free operator services, and voice calls “over all distances”—come straight from the POTS era. Modern wireless, cable, and voice-over-internet-protocol (VoIP) providers need not bundle those features into their offerings.
Worse, the California Public Utilities Commission (CPUC) has refused to let a carrier satisfy its COLR obligation with any substitute technology unless the agency first adopts service-quality standards for that substitute. For mobile wireless, it never has. The practical effect is that AT&T can meet its COLR duty only by continuing to operate the same copper network the FCC would let it retire.
Nor is there an easy exit. CPUC rules allow a carrier to relinquish COLR status only if another carrier volunteers to take it on. When more than 200 carriers were offered the chance to assume AT&T’s duties, every one declined. Shocking, yes: No firm wanted an unfunded mandate to run obsolete infrastructure.
When AT&T applied in 2023 to relinquish the designation, the CPUC spent more than a year on contested proceedings, discovery, and statewide hearings. It then dismissed the application on a threshold motion and barred AT&T from reapplying for a year.
California also layers on procedural burdens that independently block incumbent carriers from retiring POTS. State law requires basic service to be tariffed—meaning offered under rates and terms filed with the regulator—and bars carriers from removing it from those tariffs. Any change requires a formal CPUC proceeding the agency can reject.
Discontinuing POTS also triggers customer-notice approvals and a 19-point “exit plan” under the state’s Mass Migration Guidelines, with service required to continue until the agency approves each submission. Each step needs separate sign-off. None carries a deadline the CPUC must honor. Any one can be denied.
The result is a practical veto over modernization—even after the FCC says yes.
The High Cost of Hanging On
It is tempting to dismiss this as California’s problem. It is not.
AT&T and other communications providers do not budget on a state-by-state basis. They allocate capital across nationwide networks, and capital is finite. Every dollar AT&T must spend maintaining aging copper loops and switching offices in California is a dollar it cannot invest in fiber or wireless infrastructure elsewhere.
The numbers are striking. AT&T spends roughly $6 billion a year—about 5% of its revenue—operating a legacy network for a customer base that shrinks every month. California’s COLR mandate locks a significant share of that spending into obsolete infrastructure. Consumers nationwide bear the opportunity cost when faster, more reliable networks arrive later, or never arrive at all.
The market case for retiring copper is overwhelming.
Consumers have already moved on. Copper last-mile subscribers fell about 81% between 2014 and 2024, from roughly 66 million to 12.5 million. By 2024, 79% of U.S. adults lived in wireless-only households, while fewer than 1% relied exclusively on landlines.
Fiber is also dramatically cheaper to operate. Verizon’s migration of 4.5 million circuits to fiber saved roughly $180 million annually and reduced maintenance dispatches by about 60%. All-fiber networks cost about $91 less per home each year than copper-based digital subscriber line, or DSL, networks. Much of those savings come from lower energy use. Altafiber, for example, reported that copper service uses about 172 kilowatt-hours per subscriber annually, compared with just 6 kilowatt-hours for fiber—a 97% reduction.
The equipment is also nearing the end of its useful life. The Lucent 5ESS, Nortel DMS-100, and Siemens EWSD switches that anchor many copper networks have not been manufactured in decades. Carriers increasingly scour secondary markets for replacement parts. In one memorable example, Tinker Air Force Base kept a 5ESS switch running by buying components on eBay.
Copper has also become a lucrative target for thieves. As copper prices climbed from about $2.29 per pound in 2020 to nearly $6 by early 2026, theft surged. AT&T reported about 8,700 theft incidents in 2025, costing roughly $76 million. Across the industry, more than 15,500 theft-and-sabotage incidents in a single year disrupted service for more than 9.5 million customers, including 911 systems, hospitals, and military bases. Requiring carriers to keep valuable copper in the ground long after customers have abandoned it effectively subsidizes organized theft.
The upside of completing the transition is equally substantial. One Brattle Group estimate projects that finishing a nationwide fiber buildout would generate roughly $3.24 trillion in net present value and support about 380,000 jobs. That estimate reflects a complete national transition, not California’s rules alone. Still, it shows the scale of the investment opportunities foreclosed when billions of dollars remain tied up maintaining obsolete infrastructure.
Most importantly, retiring copper does not leave consumers without telephone service. Voice communications have undergone the kind of creative destruction that characterizes competitive markets: newer, better technologies have displaced the switched-access copper network, and consumers have overwhelmingly embraced them.
The FCC’s own data tell the story. Hundreds of millions of mobile subscriptions and tens of millions of interconnected VoIP lines now dwarf a residential switched-access network that continues to shrink at double-digit annual rates.
These alternatives do more than replicate POTS. They offer faster speeds, lower latency, greater reliability, and the ability to combine voice, video, and messaging in a single internet-based service. Recognizing that reality, the FCC’s 2026 Network Modernization Order permits carriers to discontinue copper service only when an adequate replacement is available. In other words, the consumer-protection function California’s COLR rules supposedly serve is already provided under federal law by the same order that authorizes carriers to retire their copper networks.
Where the Wires Cross
That leaves a legal collision.
The FCC’s 2026 Network Modernization Order streamlined the process for retiring legacy networks and expressly preempted state laws that prevent carriers from discontinuing interstate and jurisdictionally mixed services. “Jurisdictionally mixed” simply means a service that carries both interstate and intrastate traffic, as modern communications networks often do.
The order also rests on a straightforward statutory command. Section 214 of the Communications Act provides that, once the FCC authorizes a discontinuance, a carrier may proceed “without securing approval other than such certificate.”
California demands precisely the additional approvals that Section 214 appears to forbid. The complication is that the FCC’s authority under Section 214 does not extend to services that are purely intrastate—communications that begin and end within one state.
In theory, that distinction is clean. In practice, modern networks have made a mess of it. Voice calls, data packets, mobile connections, and VoIP services do not always respect the tidy jurisdictional boxes designed for a simpler telephone system. Whether the FCC can preempt California’s COLR regime therefore turns on the nature of the service at issue.
Even when a state insists it is regulating only intrastate service, federal law recognizes an important limit. Under the long-established “impossibility exception,” courts have held that federal law may preempt state regulation when interstate and intrastate components cannot realistically be separated—not only because they are physically intertwined, but because separating them would be practically or economically impossible.
That is the question the FCC now faces. If California’s COLR rules operate as a backdoor veto over federally approved discontinuance of interstate or mixed services, federal law should control. States may protect consumers. They may not use yesterday’s consumer-protection rules to override a federal decision that yesterday’s network can finally be retired.
The Last Call for Copper
California’s COLR rules were designed to protect consumers during the transition from monopoly telephone service to competitive markets. That transition happened years ago. What remains is a permanent tax on modernization—one that singles out a single provider, protects few consumers whom the market has not already reached, and quietly reduces the capital available to build fiber and wireless networks across the country.
The FCC likely has the legal authority to end this stalemate. The economic case for doing so is even stronger.
When the federal government has determined that a century-old network can finally be retired, no single state should be able to force it to live forever. At some point, even the last copper wire deserves to be disconnected.
