Splitting a national monopoly into regional ones produces the worst of both worlds

When a national network monopoly is broken into geographic pieces, the resulting regional operators don’t compete with each other (customers can’t switch regions) AND they lose the economies of scale and coordination capacity of the unified system.

Explanandum

Why did the AT&T breakup and UK rail privatisation both produce outcomes widely regarded as unsatisfactory, despite being motivated by pro-competition reasoning?

Substance

The 1984 AT&T breakup created seven Regional Bell Operating Companies, each with a geographic monopoly over local telephone service. A customer in New York couldn’t switch to the California Bell company. The competitive element was limited to long-distance service. The Baby Bells then spent two decades reconsolidating — by the mid-2000s, seven regional monopolies had become essentially two (AT&T and Verizon).

UK rail privatisation (1993-94) was arguably worse: it split the system into functional layers (Railtrack for infrastructure, 25 Train Operating Companies for services, ROSCOs for rolling stock) that had to interoperate but couldn’t compete. No passenger could choose between operators on the same route. The structure produced coordination failures, misaligned incentives (contributing to the Hatfield crash of 2000), and rent extraction through the ROSCO leasing structure. The trajectory since has been toward reconsolidation under Great British Railways.

In both cases, the network character of the good meant that fragmentation destroyed coordination value without creating competitive discipline. The market’s own revealed preference — reconsolidation — confirmed that the breakup structure was unstable.

Supports

• Baby Bells reconsolidated into duopoly within 20 years
• UK rail is being reconsolidated under Great British Railways
• The Victorian railway precedent showed the same pattern: competition → waste → consolidation → cartel/monopoly
• Neither the Olley-Pakes decomposition nor conventional metrics would detect this structural problem

Challenges

• Long-distance telephone competition (AT&T vs MCI vs Sprint) was genuine and did benefit consumers
• Some franchise competition at the bidding stage may have extracted better terms than a pure public monopoly
• The problems may reflect poor regulatory design rather than an inherent flaw in the breakup concept

Open Questions

• Is there a way to break up a network monopoly that preserves coordination while introducing genuine competitive pressure?
• Does the pattern extend to digital platform monopolies — would breaking up Google or Meta into regional versions produce the same pathology?

Source Context

Emerged from linking Albrecht’s AT&T case to the UK rail privatisation experience, observing that both followed the same pattern of splitting a national monopoly into non-competing regional pieces.