Competition and subsidies in the deregulated US local telephone industry

• Author: Mo Xiao,Ying Fan
• Published date: 01 October 2015
• DOI: http://doi.org/10.1111/1756-2171.12109
• Date: 01 October 2015

RAND Journal of Economics

Vol.46, No. 4, Winter 2015

pp. 751–776

Competition and subsidies in the deregulated

US local telephone industry

Ying Fan∗

and

Mo Xiao∗∗

The 1996 Telecommunications Act opened the monopolistic US local telephone industry to

new entrants. However, substantial entry costs have prevented some markets from becoming

competitive. We study various subsidy policies designed to encourage entry.We estimate a dynamic

entry game using data on potential and actual entrants, allowing forheterogeneous option values

of waiting. We find that subsidies to smaller markets are more cost effective in reducingmonopoly

markets, but subsidies to only lower-cost firms are less cost effective than a nondiscriminatory

policy. Subsidies in only early periods reduce the option value of waiting and accelerate the

arrival of competition.

1. Introduction

Many telecommunication services have been deregulated in the last few decades, including

the US local telephone industry. Before deregulation, services were provided by regulated mo-

nopolists, competitive entry was forbidden, and prices were set by federal and state authorities

according to cost-plus, rate-of-return regulation guidelines (Hausman and Taylor, 2012). After

deregulation, the markets were opened to competition and many of the pricing regulations were

phased out. Consequently, on the one hand, smaller, competitive telecommunications compa-

nies were allowed to enter, even using the incumbents’ unbundled network and facilities; on the

other hand, incumbents enjoyed newfound freedom and greater market power if new entrants

did not arrive. Given the substantial cost of entry in the telecommunications industry, ensuring a

competitive market structure after deregulation is an ongoing concern for policy makers.

∗University of Michigan; yingfan@umich.edu.

∗∗University of Arizona; mxiao@eller.arizona.edu.

We thank Daniel Ackerberg, Steven Berr y, Juan Esteban Carranza, Gautam Gowrisankaran, Paul Grieco, Philip Haile,

Taylor Jaworski, Kai-Uwe K¨

uhn, Francine Lafontaine, Ariel Pakes, Mark Roberts, Marc Rysman, Gustavo Vicentini,

Jianjun Wu, Daniel Yi Xu, three anonymousreferees, and par ticipants of California Institute of Technology, the Federal

Trade Commission, Harvard University, IIOC 2011, the Pennsylvania State University, SED 2012, the University of

Alberta, the University of D¨

usseldorf, the University of Michigan, and Wayne State University for their constructive

comments. Wethank the NET Institute for financial suppor t.

C2015, The RAND Corporation. 751

752 / THE RAND JOURNAL OF ECONOMICS

In general, entry costs can hinder competition in a deregulated market. When the costs of

entry are high enough, deregulation itself may not be sufficientto attract entr y.The monopoly mar-

ket structure from before the deregulation might remain, only now the incumbent is unregulated

and may exploit its market power. To address this issue, one policy remedy could be to subsidize

new firms’ entry costs. Such subsidy policies have been adopted in many industries.1This opens

up the question of how to design such subsidies as a function of the economic environment. For

instance, different potential entrants may face different levelsof entr y costs. Also, markets differ

in size, which affects post entry profit. How important is it to consider such firm heterogeneity

and market heterogeneity in the design of the subsidy policy? In addition, although a subsidy

lowers a firm’s entry cost today, it also changes this firm’s belief about the future competition

level in the market it considers entering. How important is this competition effect for the design

of a subsidy policy?

In this article, we address the above questions by estimating a dynamic oligopoly game of

entry into the US local telephone industry. Prior to 1996, local markets were served by regulated

monopolists, the so-called Incumbent Local Exchange Carriers (ILECs), who were mostly Baby

Bell companies. After the 1996 Telecommunications Act (henceforth, the Act), the markets were

opened to new entrants, referred to as Competitive Local Exchange Carriers (CLECs). In this

study,we focus on facilities-based CLECs, which build their own fiber-optic networks and digital

switches2and are deemed by industry experts to represent true competition to ILECs (Crandell,

2001, 2005; Economides, 1999).3

We use a comprehensive panel data set, which records all facilities-based CLECs’ entry

decisions into local telephone markets between 1998 and 2002. With this data set, we observe

the identity of CLECs providing local telephone services to each local market each year. We also

observe the set of CLECs with certification to enter in each state each year. In this industry, a

CLEC needs to obtain certification from a state to operate in a market within the state. After

receiving state certification, a CLEC may wait years to actually enter. Based on this industry

feature, we define potential entrants into a local market as CLECs with certification from the

respective state. With information on the identity of potential and actual entrants, we are able

to observe how long a potential entrant waits to enter a market and several crucial firm-level

attributes associated with the cost of entry.

We set up a dynamic oligopoly game and incor porate the timing of entry and firm hetero-

geneity in the game. In our model, a potential entrant is a long-run player that decides whether

to enter or wait in each period. When making this decision, the potential entrant compares the

value of entry, minus entry costs, to the value of waiting. This is in contrast to most other en-

try studies, in which a firm either enters or perishes and the value of waiting is set to zero.4

Moreover, we allow potential entrants to be heterogeneous in entry costs. For example, a more

experienced potential entrant may face lower entry costs. To estimate our model, we follow the

recent development in two-step estimation strategies for dynamic oligopolyentr y games. That is,

we first obtain the conditional choice probabilities at each state from the data. We then match the

empirical conditional choice probability with its counterpart predicted by the model.

The estimation of the model gives results that are consistent with basic economic intuition.

For instance, we find that a CLEC’s post entry profit is decreasing in competition and increasing

1This practice is especially common in service industries, where the service is considered “essential for the basic

well-being of consumers.” Forexample, from 1978 to the present day, federal governmentprograms have subsidized entry

of dentists, physicians, and mental health specialists into geographic areas designated as Health Professional Shortage

Areas (Dunne, Klimek, Roberts, and Xu, 2013).

2Facilities-basedCLECs also lease some networks from ILECs to locations not served by the CLECs’ own networks;

and, more importantly,they need to interconnect with ILECs’ networks to exchange voice and data traffic.

3CLECs that resell ILECs’ service or CLECs that rent ILECs’ networks and provide value-added services only

yield thin profit margins. They are considered as unsustainable (Crandall, 2001).

4One exception is Fan (2006), who makes the distinction between a short-run and long-run player in a dynamic

entry game.

C

The RAND Corporation 2015.