purple motes

The break-up of AT&T in 1984 is a well-recognized, major structural change in the U.S. telephone industry.  In the 1984 divestiture, AT&T’s local telephone business was separated into seven independent Regional Bell Operating Companies (RBOCs).  By year-end 2006, recombinations of RBOCs had left as AT&T successor companies the new AT&T, Verizon, and Qwest.  AT&T, Verizon, and Qwest served 46%, 32%, and 9% of subscriber lines served by the 35 largest local exchange telephone companies at year-end 2006.

A less appreciated structural change is that telephone line share and industry concentration has increased among non-AT&T local telephone companies.  From 1916 to 1942 among the 35 largest local telephone companies, the share of telephones that non-AT&T companies served fell from 13% to 8%.  However, that trend subsequently reversed. From 1942 to 2006, the telephone line share of non-AT&T companies rose back to 13%.[1]  This increase in non-AT&T share occurred despite an AT&T successor acquiring the local telephone business of Southern New England Telephone.[2]  AT&T successor companies’ spinning off rural exchanges and less reduction in rural telephone line counts helps to explain the five percentage point growth in non-AT&T telephone share.

Concentration has increased among non-AT&T local telephone companies.  The telephone share of the top-5 non-AT&T local telephone companies among the top-35 non-AT&T local telephone companies rose from 47% to 63% to 78% from 1916 to 1942 to 2006.  The five largest non-AT&T local telephone companies at year-end 2006 were Embarq, Windstream, CenturyTel, Citizens Communications, and Cincinnati Bell, with 35%, 16%, 11%, 11%, and 4% of non-AT&T top-35 subscriber lines, respectively.

On July 1, 2009, CenturyTel acquired Embarq.  In April, 2010, CenturyTel announced that it was acquiring Qwest.  Non-AT&T local telephone companies have been a small but vibrant part of the U.S. telephone industry.

* * * * *

Data: holding companies for U.S. local telephone operating businesses, 1916, 1942 and 2006 (Excel version).

Notes:

[1] Data availability for 2006 limits the comparison to top-35 companies. For 1916 and 1942, the figures are for shares of telephones.  Telephone counts are not available after 1984.  For 2006, the figures are for share of subscriber lines (loops).  One subscriber line can serve more than one telephone.  However, the difference between shares of telephones and shares of subscriber lines is unlikely to affect significantly observed trends.  Compared to the above shares of telephones, corresponding shares of operating revenue are a few percentage points higher. But the trends are similar.

[2] SBC Communications, formerly Southwestern Bell Telephone, acquired SNET on Oct. 26, 1998.  SBC subsequently became part of the new AT&T.

Growth in the U.S. local telephone business has been strongly biased toward the largest companies getting larger.  From 1916 to 1942, only about a third of U.S. households had telephone service.  Hence opportunities for extensive growth in the local telephone business were relatively good.  Nonetheless, AT&T’s share of operating revenue among the 75 largest telephone companies rose from 89% to 93% from 1916 to 1942.  The growth bias toward larger-sized companies was not confined to AT&T.  The total operating revenue of the top 5 companies excluding AT&T, relative  to the total operating revenue of the top 75 companies, excluding AT&T, rose from 43% to 63%.  The number of companies with greater than $50,000 in operating revenue in inflation-adjusted 1942 dollars fell from 224 in 1916 to 73 in 1942.[1]  Roughly speaking, AT&T absorbed about half that revenue concentration, and non-AT&T companies the other half.  The bias towards bigness seems not to have been an effect just of AT&T’s market power.

Large telephone companies had complex corporate structures.  In 1916, AT&T controlled the New York Telephone Company, which controlled the Bell Telephone Company of Pennsylvania, which controlled the Chesapeake & Potomac Telephone Co., which controlled the Chesapeake & Potomac Telephone Co. of Virginia, which controlled the Staunton Mutual Telephone Co.  Large non-AT&T companies had similar corporate complexity.  One of the largest non-AT&T companies in 1942 was the General Telephone Corporation.  It encompassed 12 operating companies formed in 1935, each by consolidating groups of smaller telephone companies.[2]  Complex corporate structure suggests relatively more importance for managerial incentives, financing, input sourcing, and regulatory political economy compared to operating economies in determining business size.

Despite the bias toward bigness, a large number of small independent telephone companies have remained in the telephone business.  In 1937, more than 40,000 small local telephone companies accounted for about 1% of total industry operating revenue.[3]  Small telephone companies seemed to have entered the business to serve specific customers that large companies did not serve.  The early telephone business apparently didn’t have cost economies of scale.  Small companies generally grew relatively slowly.  When they failed, larger companies acquired them.

*  *  *  *  *

Data: structural change in the U.S. telephone industry, 1916 to 1942 (Excel version); fuller dataset of telephone operating companies in 1942

Notes:

[1] The statistics for 1942 include many fewer very small telephone companies than do the statistics for 1916.  This is largely an artifact of the data reporting.  Reporting in 1916 and 1942 among telephone companies with greater than $50,000 in annual operating revenue (Class A and Class B companies) was probably universal.  Reporting for smaller companies is much less comprehensive, especially in 1942. The telephone censuses provide relevant comparative data.

[2] General Telephone subsequently became GTE.  The new Bell Atlantic (a combination of the old Bell Atlantic and Nynex) acquired GTE in 2000, and became Verizon.

[3]  The number of independently owned local telephone business in the U.S. today is probably about 800. See the NTS Cost Dataset.