Thursday, February 17, 2005

Reading :: The Fall of the Bell System

Originally posted: Thu, 17 Feb 2005 12:05:28

The Fall of the Bell System : A Study in Prices and Politics

by Peter Temin

In the Preface of this book, Peter Temin thanks AT&T for commissioning this historical study and providing complete access. The relationship with AT&T shows -- The Fall of the Bell System is rather sympathetic to AT&T's position throughout -- but nevertheless the book is painstakingly researched and documented, and provides a surprising level of detail into what led to Ma Bell's breakup.

I read the book primarily with an eye toward universal service, and in that respect, the book provided some startling insights. In an earlier review, I discussed Milton Mueller's claim that the term "universal service," though it has been attributed with a unitary, fixed meaning, has changed distinctly over the years. In the course of his argument, Mueller claimed that the second definition of universal service -- 100% market penetration brought about in part by low prices -- was made possible by cross-subsidies. That is, Bell charged far more for long distance than it actually cost, then used the excess revenue to support local service. Okay. But Temin takes a different tack. He assumes that "universal service" has a fixed meaning, then goes about carefully showing that "cross-subsidies" had several different definitions!

In other words, it looks like both authors have nailed down one term in order to show the vagaries of the other.

Let's talk a bit about universal service first. Temin portrays the Bell System as a quasi-public utility run as a regulated monopoly: "Faithful to its role as steward of the nation's telephone network, the Bell System had worked for close to a century to keep ahead of the public's demand for service" (p.3). Part of this stewardship role was universal service, and Temin explains this concept succinctly, compressing two definitions of universal service and attributing the combination to Theodore Vail: "The Bell System had attained the goal stated early in the century by AT&T's legendary president, Theodore Vail: to provide every household with a telephone connected to every other telephone. 'Universal service' having been achieved, it was not at all clear where the System should go" (p.6; see also p.16). Twenty years after Vail's proclamation of universal service as a goal, Walter Gifford proclaimed that AT&T had an "unusual obligation" to provide adequate and dependable service (p.16) -- an indication that even though universal service may have been rearticulated in the 1970s, the rearticulation was not as radical nor as calculated as Mueller suggested. Indeed, Temin provides considerable evidence to suggest that the 1970s rearticulation was consistent with core values that Bell had developed in the intervening years.

The stage for the rearticulation of universal service was set by Vail's reorganization of the Bell Operating Companies, a reorganization that lasted from 1909 until (depending on how you count it) 1972 or 1979.

The companies were organized along functional lines, with three main departments: Plant, Traffic, and Commercial. Since each department contained people with similar skills and training, this style of organization facilitated internal communication and promoted operational efficiency. The similarity of format allowed communication between the General Departments at AT&T and the operating companies and between the companies themselves. (p.18)

Notice the focus on internal communication and operational efficiency, which are appropriate for Vail's strategy of becoming a regulated monopoly. This was consistent with the logic of the age, in which the rationalization of industry was seen as promoting efficiency, and efficiency was seen as an unmitigated good (pp.343-344). (Part of Temin's problem, I think, is that he accepts the logic of efficiency so fully that he evaluates deregulation in terms of how it irrationally fails to meet the goal of efficiency.) Vail's reorganization arguably outlasted its usefulness: it positioned AT&T well for the task of taking over phone service, but it did not serve well when long distance competition surfaced in the late 1950s and especially during the bad years of the 1970s. After deregulation, this structure had to give way to one of deep interpenetration -- in which workers had to be in close contact and collaboration with their counterparts at competing companies.

At any rate, the organization Vail put into place lasted until 1972, when AT&T was reorganized by markets (p.79). But that reorganization left AT&T's core values and traditions intact (p. 150, 162), so AT&T was re-reorganized in 1979 (p.164), reorienting to the business market.

Between these reorganizations, of course, a lot of other things were happening. For instance, during World War II, the use of long distance skyrocketed:

Between 1939 and 1942, the number of telephone messages carried by Long Lones had shot up from 60 to 114 million. Long Lines' profits naturally had risen too, and the FCC had asked AT&T to show cause why its interstate rates should not be reduced. Walter Gifford promptly and indignantly responded that the Bell System was engaged with the government in fighting the war. AT&T was struggling to handle the increased volume of war-related traffic; it was advertising to convince the public to call less often. Lower rates would encourage people to call more frequently. "I fail to see," Gifford concluded, "how by any stretch of the imagination a reduction in long distance rates will help win the war." (p.20)

So how could the FCC reduce long distance profits without reducing the rates? The result of this short-term decision reverberated across the following decades. For the FCC decided it was worthwhile to separate calls between interstate and intrastate jurisdictions so the rates of return could be separated. The question of how to do this was not obvious. One could conceptualize the network in two ways. In the "station to station" conceptualization, the toll call goes from one phone all the way to the other; everything in the middle, every switch and every line, collectively constitutes the toll call. In the "board to board" conceptualization, the toll call is only the portion of the call that goes from one switch to another; from the switch to the phone is considered to be local service on either end (p.20).

Since the Bell System was organized as a single long distance carrier (Long Lines) that connected local operating companies -- which were distinct in legal and regulatory status -- it was easier to implement, calculate, and conceptualize board-to-board charges. Station-to-station charges would be far harder to calculate, since "some way has to be found to separate the portion of the costs of the local exchange that can be related to toll calls from the costs resulting from local exchange service" -- that is, if you use the station-to-station method, you have to figure out how to distinguish true local service from phone-to-local-switch long distance-related service (p.21). The problem came to a head in the 1930 Supreme Court decision Smith v. Illinois Bell, in which the Court declared that the station-to-station conceptualization was the standard for determining if rates had been set too low (p.22). This decision became the basis for the FCC's solution to their wartime problem: In 1942, the FCC required AT&T to use station-to-station accounting. Doing so would add access charges from the local company to the long distance service, raising long distance rates (and constraining wartime demand) while denying AT&T increased profits (p.22). At the same time, "local rates would fall, encouraging universal service" (p.23). AT&T acceeded, although it still used board-to-board accounting internally (p.23). Separations -- and cross-subsidies -- were born out of this wartime solution.

But cross-subsidies were hard to define. Under AT&T's internal board-to-board accounting, "any division of revenue is a cross-subsidy flowing from the interstate (toll) jurisdiction to the intrastate (local) jurisdiction" (pp.24-25). But the FCC tried to define the term "cross-subsidization" in 1950 as deviations from subscriber line usage (SLU) -- a simple-sounding definition that in practice was arbitrary, varying with changes in usage (pp.24-25).

At the same time, political pressure to keep local rates low led the FCC to further increase the cross-subsidy from long distance to local service -- as technological innovations simultaneously lowered the operating costs of long distance. A large imbalance -- which is to say, a large cross-subsidy -- began to form (p.25). Costs and prices diverged enormously, but the divergence was held together because all services were under the same regulated monopoly (p.27).

This situation began to change as early as the mid 1950s. Competition started in one of the smallest telecommunications niches: private lines. Essentially, private lines are off the network -- they don't conform to the first meaning of universal service. Imagine a private line between offices of the same business, or a line between a bank and its ATM, or the line between the Batcave and Commissioner Gordon's office; one phone is "hardwired" to the other. Plenty of communications-oriented businesses, like the TV networks, had the skill, equipment, and motivation to set up their own private communications using microwave systems (p.28). Motorola became the primary advotate for this arrangement, and in 1959 the FCC allowed them and others to compete in this small market niche. The Above 890 decision, as small as it seemed at the time, was the beginning of the end of the monopoly (p.29).

AT&T's argument against the decision was that it "would encourage 'cream skimming,' that is, the intrusion of independent operations in the most profitable parts of the telephone network" (p.30). AT&T's price averaging had made the price of interstate calls uniform across the country -- averaging, that is, provided another cross-subsidy -- and this monopoly logic was threatened when the most profitable routes could be taken over by competitors, since AT&T would have to raise its rates across the board to compensate for the loss of a profitable route (p.30). Furthermore, unlike AT&T's uniform level of service across the network, competitors could offer lower-quality services -- again meaning that the low-cost, high-profit customers wouldn't be around to subsidize the other customers, resulting in price increases across the board (p.30). These prices were the result of historical as well as geographical averaging, which is to say that price averaging served to pay for aggregated costs in installing infrastructure as well as geographical disparities in providing current service (p.34). The FCC found this argument convincing enough to ban the sharing of microwave systems, but it proceeded with the Above 890 decision, reasoning that it would affect only a tiny fraction of the market. That may have been true at the time, but it fundamentally changed the industry's logic and lay the ground for later competition.

At about the same time, other things were happening that would have a similarly large impact. Hush-a-Phone Corp. v. U.S. (1956) and the FCC's Carterfone decision (1968) together demolished AT&T's longstanding argument that anything connected to its network could be harmful and should be subject to tariff. The hush-a-phone was a small plastic cup that attached to the phone's mouthpiece; the Carterfone was a device that allowed you to relay a phone call from a handset over a radiophone. Neither presented a physical threat to the network, but both presented legal threats; when these two cases were decided in favor of AT&T's competitors, an important wall to competition was eroded (p.43). AT&T tried to salvage the network's integrity by allowing connections through a protective coupling arrangement (PCA), a small device that would fit between the network and the foreign device, providing a clear delineation of where AT&T's network (and responsibility) ended (p.46); but the PCA was later seen as an unreasonable attempt to preserve the monopoly.

When you put together the Above 890 decision, the Hush-a-Phone decision, and the Carterfone decision, you get ... MCI. In 1963, Microwave Communications, Inc. applied for permission to build a private line between St. Louis and Chicago, with the intent to sell capacity to others. This skirted the Above 890 decision, which barred the sharing of facilities, because MCI wasn't sharing facilities -- it was just selling capacity (p.47). MCI framed its service as new, not competitive (p.49); it escaped the nationwide averaging because of this, and it managed to keep costs low by using lower-quality construction, offering lower-quality products, and hiring nonunion laborers (p.50). The FCC accepted MCI's arguments on a 4 to 3 vote (p.50). The consequences would be grave. MCI continued over the years to encroach, until it was in direct competition for conventional long distance -- as we'll see in a moment.

In the meantime, AT&T still saw itself as having a "holy mission of service," according to Temin's sympathetic reading (p.59). Temin argues passionately here that AT&T kept prices low and falling during its monopoly, assuming that those rates would rise with the loss of efficiency that comes with competition (p.59). (In fact, prices have continued to drop rapidly in the post-monopoly era). Temin's implicit argument here and throughout is similar to Mussolini's argument that he made the trains run on time: efficiency is more prized than decentralized competition.

At any rate, by 1970 "universal service had been achieved" (p.62), at least in terms of the first and second iterations of the term. The US had reached over 90% market penetration. So the goal turned from universal service to continuous enhancement of the switched network (p.62). Also at about this time, the network began to take on an abstract identity separate from the Bell System (p.64). By 1972, the FCC gave "permission for people to connect terminal equipment to the telephone network at will" as long as they registered their equipment with the FCC (p.64). (This opened the door for innovations from answering machines to cordless phones.)

These changes roiled the Bell System. When John deButts took over as CEO of AT&T in 1972, he decided to change the focus from technology and network design to the relationship between AT&T and the regional Bell operating companies (RBOCs) (p.71). Along these lines, he began to articulate universal service as high market penetration coupled with low prices (p.72), a tack that allowed him to argue for a regulated monopoly and against cream-skimming by AT&T's competitors. deButts' new focus did not go down well with the FCC.

As a series of proposed tariffs were shot down by the FCC, and competition with MCI began to heat up, AT&T underwent restructuring. The company would now be organized by markets, not operations (p.80). And in congruence with this new orientation, "No longer was the enterprise's central mandate to supply plain old telephone service, and to pursue the goal of universal service. The need was to supply more varied demands than before, to differentiate products for particular customers" (pp.81-82). The problem with this new orientation was that AT&T was not willing to give an inch: it tried to erect its own Maginot Line, competing vigorously in all existing and emerging telecommunications markets, and refusing to let go of the integrated structure that had worked so well in the early days of the monopoly (p.83).

deButts' 1973 talk to the National Association of Regulatory Utility Commissioners (NARUC), entitled "An Unusual Obligation" (a reference to Walter Gifford's 1927 NARUC address), was meant "to defend the public switched network against competition" (p.96). The speech's aggressive positioning was well received within the Bell System, affirming as it did the traditional value of universal service (in its second articulation; see also p.144, where Temin clearly argues that this articulation reflected AT&T's core values in the 1970s). But outside the company it was another story. The speech attracted the attention of those who worried about the company's antitrust status, including "the FCC, the Defense Department, the Senate Judiciary Committee, and the Justice Department" (p.100). At the same time, MCI continued to test the limits of competition, moving to interconnect with Bell's network and positioning in legal, regulatory, and technological terms to expand into long distance service (p.103). But the real trouble began when President Nixon resigned in 1973 (p.108): the weakened executive branch could no longer restrain the Department of Justice, which had long been unhappy with the compromises made on AT&T's behalf. In particular, DoJ wanted to vacate the Consent Decree of 1956, reopening a monopoly suit.

Furthermore, tensions among AT&T's roles were becoming unbearable:

The FCC regarded MCI as a customer of AT&T and was applying regulatory doctrines of nondiscrimination to the Bell System's responses. The Justice Department regarded MCI as a competitor and was applying antitrust standards to these same actions. Both of them, furthermore, regarded AT&T's reactions to the new demands of the network as too aggressive and ungenerous. What appeared to the Bell System to be protective of the network appeared to the Justice Department to be protective instead of AT&T's monopoly. (p.109)

The result: The Department of Justice filed suit against AT&T in 1974 (p.110).

All sorts of maneuvering ensued. AT&T supported the Consumer Communications Reform Act (CCRA), which instantly became known as the "Bell Bill" (pp.118-119). CCRA "reaffirmed the nation's commitment to universal service" (p.119 -- the second articulation), and "bluntly asserted that the existing rate structure, by which it meant primarily separations, had promoted universal service" (p.119). What's more, we start to see hints of the third articulation of universal service, the continual moving of the goalposts: in deButts' testimony to a Congressional subcommittee, he described AT&T's future role as continually developing new communication technologies to add new and better services at lower and lower prices (pp.122-123). The impetus for the current articulation of universal service had begun.

Now we get back to cross-subsidies. Temin complains that the term never really gets nailed down in a satisfactory way. The FCC defined them in one way in 1950, then revised its definition in 1976, meaning that interstate tariffs that showed no cross-subsidy in 1975 would show one in 1976 (p.126)! These were in addition to AT&T's board-to-board definition used before World War II (p.127). To deal with some of this mess, AT&T began to call board-to-board cross-subsidies "contributions" (p.127).

At about the same time, MCI began to offer its Execunet service in 1975. Billed as a private line service, Execunet was gradually revealed as message toll service -- that is, it was connected on either end to the switched network. MCI was offering long distance! And the services were revealed so gradually that the FCC didn't realize it until an AT&T VP went to the Commission's offices and used Execunet to place a long distance call to a recorded weather service in Chicago (pp.132-133)! Now the competition was truly joined. The FCC ruled against MCI, but was overruled in court (p.134). By 1979, AT&T had negotiated a deal with MCI, AT&T, Sprint, and others to use AT&T's Long Lines at a discount (p.140). This deal kept the wolves at bay for a while, allowing AT&T to focus on the other problem: The DoJ lawsuit.

That problem was more intractable. AT&T's lawyers argued in 1978 "that AT&T's traditional corporate goals -- universal service and network optimization -- could not be pursued in a fragmented, competitive market" (p.166). Charles Brown, AT&T's president, re-restructured to complete AT&T's orientation to the market, focusing more on business rather than residential customers. To put it another way, cream skimming was not just acceptable but necessary in a competitive environment; the universal service orientation of a public utility had to give way to the business orientation of a competitive market (p.168). Although deButts was still committed to service as a primary goal and still thought of AT&T as a public utility, he recognized the necessity of the new orientation (p.172). And when Brown took over as CEO in 1979, he completed the turn. He identified three problems:

- First, the cross-subsidies problem: AT&T had to find a way "to balance the need for competition and the need for universal service."

- Second, the tension between competition and an efficient national network.

- Third, the basic character of the industry: was it a public utility or open competition? (p.175)

The first problem became thornier as the 96th Congress (1979) introduced a bill to, among other things, prevent cross-subsidies -- but which destabilized the meaning of the term "cross-subsidies" even more (pp. 179-180). (Temin later notes that "cross-subsidies" eventually stood for all of the industry's problems with competition (p.244).) At about the same time, the FCC shifted its rationale from the goal of lower costs to the goal of competition for its own sake (p.195) -- a rationale under which AT&T's traditionally strong case for monopoly could not gain any more traction.

Finally, AT&T saw the writing on the wall. The case was not going well, the rationale of competition did not provide criteria under which it could win, and the best Brown could hope for was a divestiture that preserved AT&T's vertical structure. In December 1981, he asked the DoJ to write a modification of final judgment for the 1956 Final Judgment (Consent Decree) (p.269). Language from a Senate bill was spliced in, and the principals (ironically) negotiated extensively over the phone before coming up with an acceptable outline for divestiture (pp.271-272). Despite resistance from the judge trying the DoJ case, resulting in several compromises, the basic agreement held. AT&T divested seven RBOCs and repositioned them to function as competitors. Ironically, the divestiture broke the Bell System along board-to-board lines; despite the rejection of board-to-board theory in Smith v. Illinois Bell, it had prevailed in a rather material way (p.308).

Let me close by pointing out a few things that are really interesting to me.

First, abstract terms such as "universal service" and "cross-subsidies" are rearticulated multiple times. Tenin tries to hold "universal service" relatively steady and Mueller tried to hold "cross-subsidies" steady so that they could interrogate the other term. But both were in flux, and often it was unclear who if anyone had the authority to fix them. Even legislators and regulators had (or exercised) relatively little power over them.

Second, the fall of the Bell System seemed to be related to a sort of tipping point being reached in the legislative and regulative culture. The Bell System's monopoly status was saved first by World War II, then by the Cold War, because the infrastructure was seen as a public utility and the logic of public utilities is rationalization leading to broad services at low prices. In this sense I think Mueller is probably wrong to argue that the second articulation of universal service was made of whole cloth in the early 1970s. Temin makes a good case that this second articulation was an expression of AT&T's core values and reflects the logic of a public utility. But it was expressed in the 1970s to counteract the new logic of competition -- not for lower prices or better service or higher penetration rates, although all three were presumed to be natural results, but for its own sake. Once the warrant of efficiency through rationalization had been traded in for the warrant of competition, the organization could not prevail. AT&T's network, which was a material instantiation of the former argument, could not support the latter argument. I think Temin understands this to some extent, but he doesn't "get" it and he consequently clings to the notion that AT&T has been wronged.

Third, there's a tremendous amount of splicing going on in the legislative battles. Some of it is quite direct: the MFJ incorporated language directly from Senate bills, for instance. Some of it is material and underhanded, such as MCI's Execunet service. Some of it is more subtle, such as AT&T restructuring its sales organization along IBM's lines. Some of it is ridiculous, like the little plastic Hush-a-Phone cups that infected the network (in a regulative sense) and made interconnections possible. But in all of these cases, junctures (and often disjunctures) created the connections that would later develop in different ways, bringing in even more actors. When the definitions of "universal service" and "cross-subsidies" slipped, when they were reinterpreted by different actors in different activities, those changes became part of the regulative, legislative, and material and practical structure of the industry.

What a long review. Time to take a break from telecommunications and read an unrelated book!

>

Blogged with Flock