A Tale of Two Networks: Early Telephony and the Commercial Internet

Daniel Mah
School of Law
Stanford University
December 2001

Thousands of networks make up the commercial Internet today. These networks are owned by many different and often competing Internet Service Providers (ISPs) that have voluntarily interconnected their networks. In the early telephone industry (1894-1921), there were also many competing telephone networks, but these companies chose not to interconnect. The result was a fragmented telephone industry that lasted until AT&T monopolized the industry and government regulation was introduced. My research explores why these otherwise similar industries evolved differently and the policy implications of my findings.

AT&T enjoyed a monopoly in telecommunications under government regulation for close to 50 years. However, in the last 30 years, US telecommunications policy has been shifting from this regulated monopoly model of service provision to a competitive and deregulated model. An important question that arises is how telecommunications firms will behave in a competitive and deregulated environment. Of particular policy concern is whether rival companies will interconnect their networks so as to create a larger (and more socially beneficial) network.

While economists have constructed stylized models of firm behavior in telecommunications industries, my thesis takes a more empirical approach. I compare and contrast two historical examples of telecommunications industries that were both competitive and not subject to direct government regulation - the early US telephone industry (1894-1921) and the commercial Internet (1994-present). Interestingly, these two industries evolved quite differently. Early telephone companies did not fully interconnect their networks, while commercial ISPs did so voluntarily. By discovering why these industries diverged, we gain a better understanding of the relationship between competition and interconnection in such industries and the effect that government actions have had on both. My findings will in turn inform future policy decisions on these issues.

The early US telephone industry and the commercial Internet are similar in many ways as both consist of networks for communication. Networks become more valuable to both existing and new users as more people join the network. Both industries also consist of many rival networks competing under minimal government regulation. However, early telephone companies did not fully interconnect their networks. In fact, a unified telephone system did not emerge until competition had been most eliminated and the regulated monopoly model was in place. In contrast, thousands of competing ISPs have voluntarily connected their networks without government regulation. Through two detailed case studies of these industries, I examine:

(1) how competition developed in each industry;
(2) why competing firms in each industry either agreed or refused to interconnect their networks;
(3) the terms of the interconnection arrangements (if any) that were employed;
(4) how such arrangements affected competition; and
(5) the role of government actions and policies in shaping the behavior of firms.

Each case study will clarify what happened in each industry, and a comparative look at why they evolved so differently will help identify the conditions under which competitors in similar industries will voluntarily interconnect their networks. In turn, this will help policymakers come up with better strategies for fostering interconnection and competition in telecommunications.

My research is not yet complete, but one interesting finding is that a competitive and interconnected Internet is not the product of "natural" market forces, as is commonly thought. Rather, my research suggests that it is unusual. The first instinct of competing firms in previous unregulated network industries is to refuse interconnection. Instead, the universal connectivity of the Internet was the result of important incentives created by the US National Science Foundation when it privatized the Internet in the early 1990s. However, my research also suggests that as the Internet becomes more commercial, universal connectivity may not persist without continued government oversight.