AT&T Dumped 7 Local Monopolies in 1984 to Keep the Profitable Half. One Bought It Back for $16 Billion and Took Its Name.

AT&T was a vertically integrated telephone monopoly that charged above-cost for long distance to subsidize below-cost local calls, used the same leverage to force its operating companies to buy only Western Electric equipment, and denied rivals like MCI the interconnection they needed to reach customers. Judge Harold Greene's 1984 remedy was structural: split the natural-monopoly local exchanges from the competitive long-distance and equipment markets, so cross-subsidy became impossible. The press called the deal an AT&T victory. Twenty-one years later, a discarded local "Baby Bell" bought AT&T for $16 billion and took its name.

On January 1, 1984, the largest company in the world dismembered itself. AT&T, with more than $67 billion in assets, over 80% of the nation's telephones, and more than 90% of interstate calls, spun its 22 local Bell Operating Companies off into seven Regional Bell Operating Companies, the "Baby Bells." AT&T kept Long Lines (long distance), Western Electric (manufacturing), and Bell Labs (research), and it was freed from a 1956 consent decree so it could finally sell computers. The local companies, which everyone agreed were the subsidized, less profitable half, walked out the door with 77% of the assets and a third of the revenue. The press reported the settlement as an AT&T win: it had kept the profitable parts and dumped the charity cases.

The mechanism the breakup was built to destroy was cross-subsidy. AT&T's leaders had long argued their cross-subsidies were a virtue: charge above-cost for long-distance, urban, and business calls, and use the surplus to keep local, rural, and residential rates below cost. That was "universal service," and AT&T said it only worked because the company held the whole pipe. Let a competitor target the overpriced long-distance routes, and the surplus would evaporate and local rates would spike. The government's view, sharpened by Assistant Attorney General William Baxter, was the opposite: the cross-subsidies were the antitrust violation. AT&T was using monopoly local revenue to underprice long distance and strangle MCI, and using its control of the local operating companies to force them to buy Western Electric, which supplied about 90% of the Bell System's equipment. The vertical integration made the abuse self-executing.

Two cracks had already opened. In 1968, the FCC's Carterfone decision struck down AT&T's tariff banning any device not made by Western Electric from attachment to the network; for the first time a customer could plug in someone else's phone, modem, or answering machine. Then MCI, allowed in 1969 to run a private microwave line between St. Louis and Chicago, sued AT&T in 1974 for blocking its expansion and won $1.8 billion in damages (later reduced on retrial to $113 million). The D.C. Circuit's Execunet rulings in 1977 and 1978 forced the FCC to let MCI compete in ordinary long distance and forced AT&T to interconnect. AT&T had fought every crack by denying interconnection, and the government concluded that conduct remedies, policing AT&T's behavior tariff by tariff, would never catch up.

So on November 20, 1974, the Justice Department filed United States v. AT&T under Section 2 of the Sherman Act. Judge Harold Greene took the case in 1978, trial began in 1981, and in January 1982 AT&T, seeing it would lose, settled. The remedy was structural, not behavioral. The seven Baby Bells were restricted to natural-monopoly exchange access only, forbidden to offer long distance, manufacture equipment, or provide information services, and required to give AT&T, MCI, and Sprint equal access on equal terms. A company that owns only the local monopoly cannot cross-subsidize a competitive business, because it has no competitive business. The abuse becomes structurally impossible, which is the point: no one can litigate the "right" internal transfer price, so you make the transfer impossible.

Then the reversal. The "loser" local monopolies turned out to own the one asset that mattered, the last-mile bottleneck, and they charged AT&T and the new long-distance carriers handsomely to reach it. AT&T's long-distance revenue share fell from 90.1% in 1984 to about 37% by 2000; its rates dropped about 30% by 1986 and kept falling. AT&T's bid to become a computer company failed. Western Electric, stripped of its captive customers, was spun off as Lucent in 1995. On November 18, 2005, SBC Communications, formerly Southwestern Bell, one of the seven discarded Baby Bells, acquired AT&T for about $16 billion and adopted the AT&T name. The thing everyone had called the winner had withered; the thing everyone had called the charity case had bought the parent.

The takeaway a modern operator should steal: when a monopolist's vertical integration lets it tax a competitive market through a bottleneck it owns, conduct remedies, regulating the price it charges rivals, almost never work, because no regulator can compute the "correct" transfer price fast enough to matter. Structural separation, making the abuse impossible by design, is the only reliable cure. But study what AT&T did to the cure: the monopoly piece you spin out keeps its monopoly, and over a generation it buys its siblings back. The 1984 breakup killed the cross-subsidy dead; it never touched the local bottleneck, which is why the Baby Bells reassembled into the AT&T and Verizon that exist today. Structural remedies work, but only for as long as the regulator stays awake.