Part 7 of Passing the Buck, a 15-part series on why we make less but pay more.
This installment is about phone and internet service, which has become the cleanest example in modern American life of what happens when a utility-style market is allowed to deregulate into an oligopoly. It is the most unavoidable monthly bill the average household pays after housing and food, and it is the one where the structural lock-in is most extreme. You cannot work without internet. You cannot get a job, schedule a doctor’s appointment, communicate with your kids’ school, or apply for unemployment without internet. You cannot function without a phone. And in most of the United States, you have one or two companies that can sell you either service, neither of which competes meaningfully with the other on price.
The structure
Three companies — Verizon, AT&T, and T-Mobile — control roughly 99 percent of U.S. wireless service. T-Mobile bought Sprint in April 2020, completing the consolidation that began with Cingular’s $41 billion acquisition of AT&T Wireless in October 2004. (Cingular’s parent, SBC Communications, renamed itself AT&T the following year after acquiring the long-distance AT&T’s name and assets.) There used to be a half-dozen national carriers. There are now three.
Two companies — Comcast (Xfinity) and Charter (Spectrum) — control roughly 60 percent of U.S. wired broadband subscribers. They do not, in most markets, compete with each other. Their service areas are carved up geographically, so a given household typically has access to one or the other but not both. Where a phone company offers fiber (AT&T Fiber in some Southern and Midwestern markets, Verizon Fios in the Northeast, smaller regional providers elsewhere), there is a second option. Where it doesn’t, there is effectively one. In roughly forty percent of U.S. zip codes, Comcast, Charter, or AT&T is the only realistic broadband provider. In rural areas the share is higher.
That duopoly is, as of three months ago, about to become a more concentrated one. On May 16, 2025, Charter announced a $34.5 billion acquisition of Cox Communications, the third-largest U.S. cable operator. The FCC, under Trump-appointed chairman Brendan Carr, approved the deal on February 27 of this year. The combined company will keep the Cox name and use Spectrum as its consumer brand, and it will be, on completion, the largest residential internet service provider in the country — roughly 38 million subscribers, passing Comcast for the top spot. The FCC’s approval order cited rural infrastructure investment commitments and a Charter pledge to onshore call-center jobs that Cox had outsourced; it also cited Charter’s agreement to end DEI programs as a public-interest benefit, which gives a sense of which considerations the agency was weighting under the new administration. State regulatory approval is still pending. The contrast with the Kroger-Albertsons grocery merger that the Biden-era FTC blocked in December 2024 — covered in Part 6 of this series — is hard to miss. Same dollar magnitude. Opposite outcome. Different administration.
I am one of the people on the wrong side of this geography. I live in the rural Hudson Valley, where the only wired option in my immediate area is the regional cable provider. Cellular signal at the house is workable but not great. There is no fiber. There is no second cable company. The price for service runs about a hundred dollars a month for an internet-only connection with the equipment-rental fee folded in. The price for the same service in most European countries, on a fiber connection at multiples of the speed, runs $30 to $40. The differential is not because rural infrastructure is more expensive in the United States — some of the costliest rural builds are in Scandinavia, where prices are still half of mine. The differential is because there is no competitor where I live, and the cable company knows it.
How we got here
There is a clean history here, and it is worth walking through, because the regulatory choices that produced the current market are unusually well-documented.
For most of the twentieth century, American telecommunications was a regulated monopoly. AT&T — the original Bell System — provided nearly all phone service in the country under what was essentially a public utility model. The company was guaranteed a monopoly. In exchange, its rates were regulated by federal and state utility commissions, and it was required to provide universal service, including to rural areas that would not have been served on a pure profit basis. Everyone had a phone. Rates were low. The infrastructure was maintained.
In 1984, a Justice Department antitrust case broke up the Bell System into seven regional “Baby Bells” plus a long-distance AT&T. The intent was competition. What happened over the following two decades was a wave of mergers that reconstituted most of the broken-up pieces into two of the three carriers that dominate wireless today — Verizon (originally Bell Atlantic plus NYNEX plus GTE) and the current AT&T (originally Southwestern Bell / SBC, which acquired the original AT&T’s name and assets in 2005 after its Cingular subsidiary had absorbed AT&T Wireless).
The 1996 Telecommunications Act, signed by Bill Clinton, deregulated nearly all of the remaining utility-style obligations on these companies. The Act’s stated purpose was to encourage competition by removing barriers to entry across telecom segments — cable companies into phone, phone companies into cable, and so on. The legislative record is full of predictions that prices would fall, service would improve, and rural areas would be connected.
What actually happened was that the deregulated incumbents merged into one another at a faster pace, built infrastructure in profitable urban and suburban markets, and left most of the rural United States with the same DSL lines they had built in the 1990s. Real per-subscriber capital expenditure by the major cable and telecom providers, as a share of revenue, has trended downward since the late 1990s. Stock buybacks and dividends are now the largest use of free cash flow for all of the major incumbents. Comcast spent more than ten billion dollars on buybacks and dividends in 2023; AT&T spent more than fifteen billion; Verizon spent more than ten billion. T-Mobile’s CEO, Mike Sievert, was paid roughly $30 million in 2024 and over $50 million in 2025 after transitioning to a vice-chairman role. Comcast’s CEO, Brian Roberts, is in a similar range. The AT&T and Verizon CEOs are in the low-to-mid twenty-million range. These are not companies that are short of money to invest. They are companies that have chosen, year after year, to return that money to shareholders rather than to the network.
A small competitive crack, for context
The one place the duopoly is showing cracks is fixed wireless. Since 2022, Verizon, T-Mobile, and AT&T have been selling home internet over their 5G cellular networks, mostly as a budget alternative to cable in markets where cable is the only wired option. They have added roughly ten million subscribers across the three carriers, almost all of them coming from Comcast and Charter. Comcast lost 181,000 broadband subscribers in the fourth quarter of 2025. Charter lost 60,000 in the first quarter of this year. The bleed is real, and it is the first time in modern memory that a meaningful share of cable broadband customers have had a working alternative.
That is the good news. The less-good news is that fixed wireless is still a duopoly product, sold by three carriers that already dominate the mobile market, with usage caps and speed limits that make it a viable alternative for many households but not for power users or small businesses. And the Charter-Cox merger that I just described above is precisely the cable industry’s defensive response to that pressure — consolidate the cable footprint, use the combined balance sheet to fund the upgrade to next-generation cable speeds, and present a more credible competitor to mobile-network home internet. Whether households end up better off depends on whether the merged Charter-Cox uses that scale to compete on price or to extract more out of the markets where it has no competition. Recent history is not encouraging on that question.
Net neutrality, briefly
Net neutrality — the principle that an internet service provider has to deliver all lawful internet content at the same priority, rather than charging websites for “fast lane” access or throttling competitors — has had a roughly twelve-year regulatory life and is now, for practical purposes, dead at the federal level.
In 2015, the Wheeler-era FCC under Obama classified broadband as a “telecommunications service” under Title II of the Communications Act, giving the agency authority to enforce net neutrality. In 2017, under Trump, FCC Chairman Ajit Pai — a former Verizon associate general counsel — repealed it. In April 2024, the Rosenworcel-era FCC under Biden restored it. In August 2024, a Sixth Circuit panel stayed the restored order. And on January 2, 2025, the same Sixth Circuit struck the order down for good, citing the Supreme Court’s 2024 Loper Bright decision, which had overturned the doctrine of judicial deference to agency interpretations of ambiguous statutes. The court held that the FCC simply lacks the statutory authority to regulate broadband as a telecommunications service, regardless of which administration is in office.
That decision means net neutrality is gone for the foreseeable future absent action by Congress, which is unlikely. ISPs can now legally prioritize, throttle, or charge for fast-lane access to specific services. They have mostly not done it openly, because the political cost would be high. But the regulatory backstop is gone.
The federal investment, late
The one piece of this story that runs in the opposite direction is the Broadband Equity, Access and Deployment program, known as BEAD, a $42.45 billion grant program created by the 2021 Infrastructure Investment and Jobs Act and aimed specifically at connecting the rural United States. The original program, as designed under the Biden administration, prioritized fiber and imposed significant labor and equity requirements on grant recipients. It was, as one industry analyst put it, “the most complicated broadband grant program ever.”
The program spent most of 2024 and 2025 stuck in administrative review, including a Trump-administration revision in 2025 that loosened the fiber-only requirement, allowed more fixed-wireless and satellite providers (including Starlink) to bid, and stripped some of the labor provisions. As of April 2026, most states have unlocked their first tranches of funding and construction is starting. The revised program is expected to reach fewer households than the original was designed for, but reaching any meaningful fraction of the rural broadband gap is a real public-policy outcome, and BEAD will land in places that the private sector spent twenty-five years declining to serve.
Whether it actually closes the gap will not be clear for several years. Fiber-supply costs are reportedly up forty percent since the program began. The Rural Digital Opportunity Fund, an earlier and smaller program, saw more than a third of award recipients default. The federal money is real. So is the difficulty of actually building the lines.
The bill, again
What that history adds up to, for the household, is a monthly bill of roughly $80 to $130 for broadband and $60 to $90 per line for cell service, sold by a company that does not face meaningful local competition, subject to fees and equipment-rental charges that are designed to make the advertised price misleading. If you have a family plan with four phone lines and home internet, the combined monthly outlay can run $250 to $350. That is a fixed cost, mostly inflexible, levied against household budgets that, as we have been documenting in this series, are already stretched against income that has not kept pace.
The international comparison is real but easy to overstate. South Korean broadband is, in fact, substantially cheaper than American broadband. So is French, German, Japanese, and most northern European broadband. The differential is not the cherry-picked twelve-times-cheaper number that gets circulated. For an apples-to-apples gigabit fiber subscription in a city, the European price is typically half to two-thirds of the American price, sometimes less. But the structural reason is the same one we have been tracing through every installment: those countries have either retained more of the utility-style regulation we deregulated in 1996, or they have a state-built fiber backbone that private companies lease access to on regulated terms, or both. The American market is unusually concentrated and unusually deregulated, and the bill reflects it.
What I see from here
I run a small business out of my house. My internet is a business expense. My phones are a business expense. The unreliable cellular signal is a business problem I have spent money trying to solve — boosters in the house, a Wi-Fi-calling fallback for when the cell tower acts up, a backup connection through my phone’s data plan for when the cable provider’s network goes out, which it does, with depressing regularity, because there is no competitive pressure for them to maintain it better than they do.
What I am paying for, in the end, is access to the most basic infrastructure of modern economic participation. It is not optional. There is no scenario in which a household opts out and remains functional. The price is set by a company that knows that.
That is the design. The cost did not go away. It moved, from the utility-style regulated rate of the pre-1996 era to the unregulated monopoly price of the post-1996 era. And the public investment that is now belatedly trying to repair the rural side of the gap is, for the first time in roughly thirty years, an actual federal program rather than a subsidized handout to the same incumbents that created the gap in the first place. That is the small, late, and welcome counterweight in this picture.
The next installment looks at insurance.
Even that's Odd 
Leave a comment