Travelling through Australia recently it was curious to observe the passionate debate over the various forms of separation which some say should be imposed on Telstra if it were to win the National Broadband Network tender.
I say “curious” because the proponents of separation keep looking to Europe and the UK at fledging separation models, or models just under consideration, instead of looking to the US where these ideas have been fully tested…and found wanting.
Over the past 25 years in the US telecommunications market, every flavour of separation has been tried, and none has been successful. I can say that having seen all sides of the argument – as a regulator following the break up of AT&T in the early ‘80s, later as a regulatory adviser inside the telecommunications company Sprint, and subsequently as an academic and teacher of regulators.
There are three basic approaches to separating competitive or potentially competitive services from apparent non-competitive operations. The most severe approach is ownership separation. The breakup of AT&T was an example of a divestiture, which included a prohibition on the carved-off regional “Baby Bells” from providing long-distance. The next most severe approach is functional/operational separation (where a company must organise its operations so that competitive market functions operate separately from the non-competitive market functions) and finally accounting separation, where a company can retain integrated functions but keep separate accounting records.
The principal problem with any form of separation is that telecommunications is a rapidly changing industry. Even if one were successful in picking the right place to draw bright lines of separation between retail and wholesale operations today, tomorrow you would be wrong.
In addition, separation is a highly interventionist regulatory measure which adds costs. The separation creates interest groups, some of whom can gain strategic advantage by changing the parameters of separation. In the US platoons of lawyers and lobbyists battled over shifting rules to benefit their particular vested interests.
And finally, empirical studies have demonstrated that separation had a measurable, negative impact on innovation - up to 50% in some instances. Complex regulatory measures made it hard to develop and take new services to market. Ironically, these well-intentioned procedures were put in place to promote innovation by helping competition, but had the opposite effect.
The AT&T divestiture was a long and drawn out lesson in how regulation can only stymie, but not contain, the broader logic of market forces. While the intent was to facilitate competition by segmenting long-distance from local-call services, consumer demand for a single calling service won out in the end, and over the course of a dozen or so years, the industry integrated again.
While the break up of AT&T is the most widely-known separation story out of the US, it is by no means the only one: The FCC ran a lengthy series of proceedings known as the Computer Inquiries, starting in the mid 60s and running about 40 years, in which it tried to draw a distinction between telecommunications and computing. The FCC was concerned that telcos might try to monopolize data processing and information services. The FCC dropped all separation requirements, but never resolved how to draw a distinction between computing and telecommunications.
The US also dabbled in a form of functional separation where incumbent telecommunications companies were forced to create separate subsidiaries to provide wholesale network services and retail services. There was a voluntary separation of Rochester Telephone in New York, and an attempted forced separation of Bell Atlantic in Pennsylvania. Both involved lengthy and diverting regulatory proceedings, and neither appeared to help competition.
The US examples should show that separation is not an end in itself, but a means by which regulators and policy makers have hoped to achieve equivalent access for competitors to networks whose owners, at the time, had no infrastructure competitors. This equivalence was ultimately achieved by the application of behavioural rules on incumbents, but in the context of reintegration, not through separation.
It emerged that requiring equivalent competitor access to the operational systems that support the network was less intrusive and more efficient than structural or functional separation. Operational support systems include automated or computer-based access to the incumbent’s ordering and provisioning systems so that access seekers can provide equivalent services to their customers.
In July, at the invitation of the ACCC, I addressed their Gold Coast conference on the topic of “what regulators need to know”. I commend the ACCC for an impressive conference. A consistent theme from nearly all of the speakers was that market forces are more effective than market planning at providing the innovations needed to compete in a global economy. Separation stifles market forces.
Dr Jamison is the director of the Public Utility Research Center at the University of Florida, and its director of Telecommunications Studies. He was in Australia in July at the invitation of the ACCC, and has also provided a paper as part of Telstra’s submission to the National Broadband Network regulatory review.