Did AT&T die in vain? An empirical comparison of AT&T and Bell Canada.(The Enduring Lessons of the Breakup of AT&T: A Twenty-Fiv

I. INTRODUCTION

Did the divestiture of AT&T in 1984 achieve its purpose of greater efficiency, lower prices, and more rapid development in American telecommunications? This question is important not only for economic historians or antitrust officials, but also for current telecommunications policymakers. In Europe, in particular, a policy of structural separation is now in the process of being mandated, (1) which aims to segment dominant and vertically integrated telecommunications companies in order to enhance competition.

Of the varieties of separation--accounting, structural, functional, geographic, and business-line--none was more radical than the AT&T divestiture. To academic economists, the divestiture was a particularly sweet moment. Their majority perspective--to create competition, if necessary, by breakup--had miraculously prevailed through the legal process against the massed political opposition of consumers, labor unions, rural folks, states, the Pentagon, Congress, and even President Ronald Reagan. (2) It suggested that the move toward competitive market structure would eventually prevail. This perspective was widespread among economists. Dissenters were either proponents of the classic public-obligation, (3) public-utility model, or advocates of theories arguing that monopoly could be efficiently contested, (4) or true laissez-faire adherents who wished competitors to enter without governmental action. (5) The first group was dismissed as behind the times, the second as beholden to AT&T, and the third as impractical purists.

Was the full structural separation successful? Did it create innovation, efficiency, consumer benefits, and investment? In terms of theory, one could argue it three ways: (1) Positively--competition is beneficial; (2) negatively--the economies of scale of a "natural monopoly" will be lost; or (3) neutrally--it makes no difference in the end because the underlying technological and economic forces are determinative. Who was right? By now, a quarter century has passed, full of impassioned regulatory and legislative battles. We have some numbers to show for this period. But how can we measure that reality against an alternate reality so we can evaluate them?

Fortunately, such an alternate reality exists--it is called Canada. Canada had a telecommunications structure very similar to that of the United States. In fact, the major Canadian carriers, Bell Canada (by far the largest company) and BC Telecom, were long owned by AT&T and GTE, as were the major telecommunications equipment makers. Canada had regional carriers along the U.S. model, closely associated with equipment manufacturers, plus many small, rural, independent companies, as well as several province-owned companies. (6) Its regulatory system was similar, with both a national and a provincial level of regulation. The major structural difference was the absence of a divestiture of Bell Canada.

II. MARKET STRUCTURE

AT&T's dominant position before the breakup was astonishing. It accounted for a full 38% of the entire media and information sector. (7) The next-largest media and information sector firm, IBM, accounted for 8.3% domestically. The second-largest telecommunications firm, GTE, had 2.2%. (8) AT&T was the world's largest private company, with over 1 million employees in the United States alone. (9) However, barely twenty years later, the company ceased to exist after successively shedding major parts and was absorbed for a mere $16 billion by its own offspring, SBC (formerly Southwestern Bell Corporation). SBC promptly renamed itself AT&T Inc., as distinguished from the historic AT&T Corp.

The gradual demise of AT&T, however, does not mean a decline of industry concentration in telecommunications. Graph 1 shows the national concentration in the overall U.S. telecommunications market. Concentration is measured by the U.S. Department of Justice (DOJ) through the so-called Herfindahl-Hirschman Index (HHI) which is the sum of the square of the percentage of market shares. The graph also depicts an alternative and more intuitive definition of market concentration, the C4 Index (C4), which is the sum of the market shares of the top four firms.

[GRAPHIC 1 OMITTED]

Graph 1 shows how the overall telecom industry has dropped in concentration from a C4 of almost 100% (of which AT&T accounted for 90.7%), to one of about 70% in 1984, after divestiture. However, after 1992, concentration rose again, consolidations were approved after significant struggle, and the C4 climbed to 85% in 2007. AT&T Inc. has a market share of 42%, while Verizon holds 28%. (11) In HHI terms, the telecommunications services industry has returned to high national concentration with an HHI of 2,986, significantly higher than the DOJ guideline threshold of 1,800 for a highly-concentrated industry, and much higher than the post-divestiture HHI of 2,145.

In 2007, Bell Canada had 45.2% of the Canadian market, including its share in Aliant, while Telus held 22%. (12) The Canadian market concentration had an HHI of 2,463. In 1984, the HHI stood at about 2,220, about the same as that in the post-divestiture United States. This is a market structure very similar to that of the United States.

Overall, the U.S. telecommunications-sector market moved, not to direct telecommunications competition, but to an oligopoly with a market structure that might be called "2.5" (two major firms, and several small ones), of parallel, regionally-dominant firms. This same trend also characterized the mobile telecommunication services industry, which was nascent in 1984, with major openings provided to new entrants. In the divestiture, AT&T lost its wireless mobile operations. In time, this sector also moved to the same market structure of two major national firms plus a few smaller operators. The top three firms, thus, are AT&T with 35.2%, Verizon with 34.5%, and Sprint Nextel with 18.9%. T-Mobile follows with 5.9% nationally, and there are several regional operators. In Canada, the market structure is similar: there are three major providers--Bell, Telus, and Rogers--each with about 30%, plus 2 larger providers in their respective provinces. (13)

Graph 2 shows the combined U.S. market shares of AT&T and its successor firms.

[GRAPHIC 2 OMITTED]

Before the divestiture, AT&T controlled almost 90% of the three classic telecommunications services: local, private line, and long distance. For local telecommunications providers, narrowly defined as phone companies in a national market, this share remained and even slightly increased by 2006. However, with a broader and more accurate definition which includes resellers, cable companies, and Voice over Internet Protocol (VoIP) providers over broadband, it has declined to 69.2%. But this did not require a divestiture.

In Canada, on the other hand, the share of incumbent telephone companies in local service dropped even more, to 62% by 2006. A move to cable TV as a major telecommunications platform can be observed in all countries where cable TV has a strong presence. Relatedly, broadband, offered by the telecommunications carriers themselves, has enabled access by independent companies offering VoIP. This, too, does not require a divestiture, and can be observed around the world. In Canada, there are 2.4 million residential VoIP and cable phone subscribers, a 50% higher penetration than in the United States. (15)

III. GROWTH

In 1983, AT&T's revenues were $89 billion (16) ($185 billion adjusted for inflation to 2007 dollars), of which $85.3 billion came from services and $3.8 billion from equipment. In 2008, all of AT&T's successor companies had combined revenues of $242 billion. (17) The revenue growth was 31% over 25 years, which is quite low. In contrast, in 1983, Bell Canada's revenues were $4.65 billion ($10.9 billion adjusted for inflation to 2007 dollars) and in 2007, $17.9 billion. (18) Bell Canada's revenue rose by 64% over the last 25 years--more than double the rate of AT&T's combined successors in the United States.

IV. PRICES

It might be objected that the slower revenue growth in the United States merely reflects a stronger price decline in the United States, and hence consumer benefits. This objection is incorrect. The Consumer Price Index for telecommunications rose in the United States from 1984 to 2004 from an index of 100 to 120. (19) In Canada, during the same period, the price index rose almost identically, from 100 to 118. (20) These were the relative price trends. The prices themselves stack up, according to the Organisation for Economic Co-operation and Development (OECD), as follows: (21) in 2007, prices in Canada were, relative to the United States, 7% higher for residential low users and 26% lower for residential high users. Prices were higher for small and medium-sized businesses (SMEs) by 40%, they were higher for business long distance by 7%, while cheaper for residential long distance by 108%. Thus, United States rates are more favorable to business users and low-use residential users, and less favorable to high-use residential users, especially those making long-distance calls. For mobile calls, the United States is cheaper, with comparable popular service baskets costing 30-67% more in Canada. (22)

V. EMPLOYMENT

U.S. telecommunications employment rose from 1990 to 2006 by 0.8%. There was actually a huge reduction in employment (25%, or 170,000 jobs) in the wireline sector, offset by growth in wireless (130,000 jobs) and cable (66,000 jobs). (23) There was a huge peak around 2001 during the Internet boom years. From 1997 to 2006, U.S. telecommunications employment declined slightly, by 1% or 10,000 jobs. In Canada, during the same period, employment increased by 20,000 jobs, from 99.5% to 119.7%, an increase of 20.3%.