**Draft**
VERTICAL SEPARATION, MONOPOLY, AND ITS CONSEQUENCES:
EVIDENCE FROM TELECOM PRIVATIZATIONS
Bruno E. Viani*
Article presented at the International Industrial Organization Conference
Boston, April 7-9, 2006
ABSTRACT
Two common policy instruments used by governments to increase the availability of
basic telephony while at the same time maintaining a low price of residential telephony
were: (1) award monopoly rights on basic services to cross-subsidize residential local
telephony; and (2) separate vertically the owner of the local fixed network from the
provider of long distance or international telephone services. Policy variation across
countries allows me to assess the effect of these policies on the use of international
telephony, local fixed telephony, and on the price of residential local telephony. I use
panel data estimation with country fixed-effects using data from 67 countries that
privatized their state-owned telephone monopoly and find that contrary to wide spread
believes: (1) monopoly on basic services is not associated with lower (subsidized) prices
of local residential service, quite the opposite, monopoly increases residential local
prices; (2) monopoly does not help universal service provision, and lowers the use of
international telephony; and (3) vertical separation does not help the expansion of
international telephony; nor that of local fixed telephony.
*
Department of Economics, Colgate University, 13 Oak Drive, Hamilton, NY 13346.
[email protected]
**Draft**
1. Introduction
Once the decision to privatize their state-owned utilities was made, governments
needed to make two important policy decisions. First, they needed to decide whether the
status quo of monopoly was to be continued in the post-privatization years; second, they
needed to decide whether the state-owned monopoly needed to be broken up
(horizontally or vertically) into different entities. The argument in favor of extending
monopoly rights in the privatization of telephone firms was political. Supporters said that
governments pursue universal telephone service in local telephony by subsidizing rates of
residential local telephony from business telephony (i.e., long distance, international, and
data communications services);1 this cross-subsidy scheme could only be maintained
under monopoly. Local fixed telephony was assumed to be a natural monopoly while
business services (i.e., long distance and international telephony) were deemed
potentially competitive. If entry were allowed on these markets competition will not
develop in local telephony while entrants on long distance and international telephony
will drive down the prices on these markets.2 An increase on local residential rates will
follow as the incumbent monopoly in local service would struggle to maintain the level of
profits it enjoyed before. Political outcry will ensue. This increase on local residential
1
Empirical evidence suggests that local residential services were priced at below marginal cost while
business services such as long distance and international telephony were priced at well above marginal cost
(Nambu, Suzuki, and Honda 1989; Crandall 1989; Palmer 1992. Hausman, Tardiff, and Belinfante 1993;
Cronin, Colleran, Miller, and Raczkowski 1997).
2
Bös (1993: 108) articulates this view clearly: “In contrast to private firms, public enterprises have often
been instructed to price according to distributional objectives. This implies charging lower prices for goods
which are mainly demanded by lower-income earners. In this case the public enterprises rely on internal
subsidization, where the internal deficit of the low-priced goods is financed by the internal profits earned
from sales to higher-income or business customers. If the privatized firm operates in a competitive market,
this internal subsidization becomes impossible and distributional pricing cannot be upheld.” Pilcher (1994:
401) puts it bluntly: “Governments, therefore, need to decide on a strategy to either introduce competition
in long-distance and international service or to maintain the cross-subsidy.”
**Draft**
2
prices would be a set back to the stated policy of universal service on local telephony.
Policy makers feared the political consequences of open entry; although, as it will
become clear later, it was not warranted.3
The validity of these arguments is suspect on several grounds. For example, it
ignores gains on productive efficiency arising from the switch to private property
ownership. Private property exerts powerful incentives to use resources efficiently
pushing down further production costs (Alchian 1965; Alchian 1969; and Furubotn and
Pejovich 1972). These gains in efficiency are compounded when considering the effect
of competition in the markets for local and long distance/international service (Alchian
and Kessel 1962; Williamson 1963; Leibenstein 1966; Comanor and Leibenstein 1969)
which will push down production costs and pass some of these gains in efficiency to
consumers in the form of lower prices ceteris paribus. Empirical studies using data from
the U.S. show that after competition was allowed on domestic and international
telephony, the corresponding increase on local residential prices did not produce a
decline on the amount of residential subscribers as it was feared (Hausman, Tardiff, and
Belinfante 1993; Cronin, Colleran, Miller and Raczkowski 1997). These results are
driven by: (1) the relatively low elasticity of demand for local access in the US; and (2)
households perceive local and long distance services as complements. Nothing ensures
that these conditions prevail in other parts of the world. This study tests some of these
arguments.
3
Although the argument for monopoly was mainly political, some analysts found an economic justification
to monopoly and cross subsidization. Faulhaber (1975) proved that cross subsidization can increase social
welfare as long as a multi-product monopolist exhibits economies of scope on joint production. Other
commentators like ___ argued that monopoly was desirable to ensure an adequate cashflow in the postprivatization years to finance the required expansion of the local fixed network ( CITE…).
**Draft**
3
A second issue policy makers needed to decide was on the desirability of
vertically separating the telephone monopoly. This time the argument was on economic
grounds and supporters came from the pro-competition side. The argument closely
followed the modification of the final judgment (MFJ) rationale used in the U.S. to justify
the split of AT&T into several local fixed telephony providers and one provider of long
distance and international service. Briefly stated, the argument assumes that competition
in local fixed telephony will not arise because of its natural monopoly characteristics. On
the other hand, long distance and international service were viewed as potentially
competitive. The supporters of vertical separation focused on how to increase
competition on this competitive segment.4 In their view, a vertically integrated monopoly
has a wide array of weapons to keep potential entrant out of the lucrative long distance
and international telephony markets. For example, the vertically integrated firm can tie
the monopolized local service with long distance and international service eliminating
competitors in the latter markets. After competitors are driven out, the integrated firm
can raise the price of the tied good reducing social welfare (Whinston 1990). Because the
vertically integrated firm owns the monopolized local fixed network it can refuse to
interconnect; delay interconnection through lengthy negotiations with potential entrants
on a fair access fee (Salop and Scheffman 1983); or it can provide low quality access
links to degrade the service quality of future competitors (Cremer, Rey, and Tirole 2000;
4
This view is clearly spelled out by Bös (1993: 108): “The government should first attempt to encourage
competition, as the UK, for instance, did with its splitting of the electricity industry into electricity
generation (a potentially competitive business) and distribution.” Also Waterschoot (1994: 511)
underscores the influence of the MFJ that split vertically AT&T in shaping the post-privatization regulation
of telecommunications around the world: “The influence of UNITED STATES ANTITRUST legislation in
shaping the structure of telecommunications cannot be questioned. Divestiture of AT&T, rather than
control over rates applied by the dominant carriers, was the main regulatory feature leading to increased
competition in long-distance telecommunications services.”
**Draft**
4
Aviram 2003). To avoid these anticompetitive behavior owners of the local fixed
network needed to be barred from providing long distance and international service.
Although these arguments seem plausible to a first approximation, they overlook an
important issue: firms integrate to minimize the transaction costs of using the market
(Coase 1937). Breaking up a firm by administrative process in the hope that social
welfare would increase should be regarded as a perilous road to follow. It implies that
the loss in efficiency after the break up more than compensates the gains in social welfare
from preventing the local monopolist to compete in the markets for long distance and
international telephony; nothing ensures this. Indeed, Williamson (1971) identified
several cases in which vertical separation may be inefficient: (1) when this creates
bilateral monopolies (i.e., double marginalization problem), (2) when bargaining
(transaction) costs between parties are likely to be high, or (3) when large sunk
investments are part of a transaction with incomplete contracts. Coincidentally, the
contracting arrangements of a vertically separated local telephone monopoly with a
downstream (international or long distance) service provider seem to have most of the
characteristics just mentioned. The key issue is that these contractual relationships
involve investment decisions that are asset specific and therefore subject to a high risk of
opportunistic behavior on either side to appropriate quasi-rents. Firms will incur high
transaction costs to contain opportunism; a way of avoiding these is to integrate (Klein,
Crawford, and Alchian 1978; Williamson 1985). When assets across firms are mutually
specific, integration increases efficiency which in turn increases social welfare
(Grossman and Hart 1986). Empirical evidence from the U.S. telephone industry
indicates that vertical separation did not produce the expected efficiency gains. Access
**Draft**
5
charges to long distance and international service providers were kept high after the break
up of AT&T to keep local residential rates subsidized; this prevented large welfare gains
from lower prices on long distance and international service (Crandall 1988; Crandall
1989; Hausman, Tardiff and Belinfante 1993).
This article tests these arguments using data from 67 countries that privatized the
former monopoly provider of basic services between 1984 and 2003. For each country I
use data for the seven-years following the privatization sale and find that awarding
monopoly rights on local fixed telephony is associated with less fixed lines in service.
Monopoly does not help universal service; on the contrary, it hurts it. As expected,
monopoly on international service is associated with significantly less usage of
international telephony. Moreover, the fears of an increase in residential telephony due
to competition did not materialize. Monopoly does not keep prices of local residential
service low (through cross subsidization) but actually it is associated with significantly
higher prices. After privatization, competition did not cause an increase of residential
local telephony, a previously subsidized service. Contrary to common believes vertical
separation is associated with significantly lower output on international service. This
suggests that the rationale behind the MFJ in the U.S. and applied in many countries was
flawed; it does not apply to the market conditions in the US, it does not apply to market
conditions elsewhere. Also important; vertical separation harms the expansion of local
fixed telephony; an issue largely overlooked in the policy debates in the pre-privatization
years.
2. The data
**Draft**
6
My dataset includes 67 countries that privatized the dominant telephone firm in
the period 1984-2003. This period covers most of the worldwide privatizations of
telephone firms. Some countries have multiple sales of blocks of equity in this period. I
identified the earliest sale as the privatization date. The main source for these data is
Privatisation International (monthly issues) and the Privatisation International Yearbook
(annual issues). Compiling the data required a page-by-page search of every monthly and
annual issue. I recorded every upcoming or completed privatization of a telephone firm.
Additional data on sales transactions were gathered from the Multilateral Investment
Guarantee Agency’s database on privatizations (Privatization Link)5 and from the
Economist Intelligence Unit’s Viewswire and Country Information databases to cover the
feasible universe of sales transactions.
Next, I collected information on the vertical structure of the telephone industry;
namely whether basic telephone services (i.e., fixed local telephony, national long
distance, or international telephony) were vertically integrated in one or more dominant
firms and the type of basic services offered by these firms. In addition I collected data on
the number of years of monopoly awarded to the privatized firms on each of the three
basic telephone services. These data were mainly collected from the Economist
Intelligence Unit Viewswire and Country Information database, the firm’s annual reports,
each country regulator’s websites, and from the Commission of the European
Communities. I also gathered information of basic telephony usage from the International
Telecommunications Union’s World Telecommunication Indicators (2004). Finally
country-wide data such as income and population was obtained from the World Bank’s
5
Available at http://www.privatizationlink.com
7
**Draft**
World Development Indicators database. Table 1 presents the full list of countries
included in my sample along with summary data on monopoly rights and vertical
separation. Table 2 provides the list of variables used in the empirical analysis along
with their respective definitions and sources.
Table 1: Summary data of privatization sales
Country
Year
Argentina
Armenia
Australia
Austria
Barbados
Belgium
Belize
Brazil
1990
1997
1997
1998
1991
1995
1988
1998
Canada
Cape Verde
Chile
Croatia
Cote D’Ivoire
Cuba
Czech Rep.
Denmark
El Salvador
Estonia
Finland
France
Ghana
Greece
Guatemala
Guinea
Guinea-Bissau
Guyana
Hungary
India
Indonesia
Ireland
Israel
1990
1995
1988
1999
1997
1994
1995
1994
1998
1993
1998
1997
1996
1996
1998
1996
1989
1991
1993
1991
1994
1996
1990
Firm sold
Telefonica and Telecom
Armentel
Telstra
Telekom Austria
Bartel
Belgacom
Belize Telecom
Embratel, Telesp, Telenorte
Leste, and Tele Centro Sul
Telus
Cabo Verde Telecom
CTC
Hravtske Telekomunikacije
CI Telecom
EMTEL Cuba
SPT Telecom
Tele Denmark
CTE
Eesti Telekom
Sonera
France Telecom
Ghana Telecom
OTE
Telgua
Sotelgui
Guine Telecom
GT&T
Matav
MTNL and VSNL
PT Indosat
Telcom Eireann
Bezeq
Separation
Vertical
yes
no
no
no
yes
no
no
yes
Monopoly a
yes
no
yes
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
yes
yes
no
no
yes
yes
no
yes
yes
yes
yes
yes
no
yes
n.a.
no
no
yes
no
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
no
no
yes
yes
yes
no
8
**Draft**
Italy
Jamaica
Japan
Jordan
Korea
Latvia
Lithuania
Malaysia
Mauritania
Mauritius
Mexico
Mongolia
Morocco
Netherlands
New Zealand
Pakistan
Panama
Peru
Poland
Portugal
Puerto Rico
Qatar
Romania
Sao Tome
Senegal
Serbia
Singapore
Slovakia
South Africa
Spain
Sri Lanka
Switzerland
Taiwan
T. & Tobago
UK
Venezuela
1997
1988
1987
1999
1993
1994
1998
1990
2000
2000
1990
1995
2000
1994
1990
1994
1997
1994
1998
1995
1992
1998
1998
1989
1997
1997
1993
2000
1997
1995
1997
1998
2000
1989
1984
1991
Telecom Italia
Telecom. of Jamaica
Nippon T&T
Jordan Telecommunications
Korea Telecom
Lattelecom
Lietuvos Telekomas
Telekom Malaysia
Mauritel
Mauritius Telecom
TELMEX
Mongolian Telecom Co.
Maroc Telecom
PTT Netherlands
Telecom of NZ
Pakistan Telecom.
Instituto Nac. de Telecom.
CPT and ENTEL
TPSA
Portugal Telecom
Tel. Larga Distancia
Qatar Public Telecom.
Romtel
CST
Sonatel
Telecom Serbia
Singapore Telecom
Slovenske Telekomunikacie
Telkom
Telefonica de Espana
Sri Lanka Telecom
Swisscom
Chunghwa Telecom
T&T Telephone Co.
British Telecom
CANTV
Year: year of first privatization sale
a If more than one year of guaranteed monopoly at the time of privatization
b merged after one year
c Only on international telephony
d only on local fixed telephony
e only on international telephony
f merged after two years
no
no
yes
no
no
no
no
no
no
no
no
no
no
no
no
no
no
yes b
no
no
yes
no
no
no
no
no
no
no
no
no
no
no
no
yes f
no
no
no
yes
no
yes
no
yes
yes
yes
no
yes
yes
no
no
yes
no
yes
yes
yes
yes c
yes
yes d
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes e
no
n.a.
yes
no
yes
9
**Draft**
Table 2: Definition of variables and sources
Variable
INTLMIN
Definition
Number of outgoing international minutes per person in year t.
FIXLINES
The country’s fixed telephone lines in service per person in year t
LOCALRES1
The average price of residential local fixed service in year t. LOCALRES1 =
Connection charge + 12 x monthly rate
The average price of residential local fixed service in year t.
LOCALRES2 = Connection charge + monthly rate/0.05
Source
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
Number of years until end of monopoly on international telephony
Own database
LOCALRES2
INTLMONOP
LOCALMONOP Number of years until end of monopoly on local fixed telephony
Own database
VERTSEP
Dummy variable if incumbent fixed line operator is different from the main
service provider of international telephony.
Own database
INCOME
Country’s real GDP per capita in PPP dollars (US$2003) in year t
MOBILE
The country’s number of cellular mobile subscriber per person in year t
WAITLINES
The Country’s number of waiting lines per person in year t
CELLP2
The average price of cellular telephony in year t.
CELLP2 = connection charge + monthly rate/0.05
Percentage of digitalization of the main fixed line network in year t.
World Bank. World
Development Indicators 2004
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
ITU. World Telecommunications
Indicators 2004
DIGITAL
ITU: International Telecommunications Union.
10
**Draft**
3. Consequences of vertical separation and monopoly on international telephony
I start testing the argument that vertical separation increases competition and
therefore output on long distance and international service. My focus is on a measure of
output rather than competition. Country data in long distance calls is limited so I use the
more readily available data on international outgoing minutes. I test the following null
hypothesis:
ƒ
H01: Separating the dominant fixed network provider form the international
telephony market increases output on international telephony.
I use panel data estimation with country fixed effects to test this hypothesis. For each
country I use eight years of data starting with the year in which the first privatization sale
took place (t0) and ending seven years later (t0 +7). I estimate the following equation:
E [Log (INTLMIN )it | X it ]= X it β i + ε it ,
(1)
where, INTLMINit is a vector of observations of outgoing international minutes per
person in country i at time t, where the initial observation for each country is at time t0
(the year of privatization), and the last observation is at time t0+7. Xit is a matrix of
exogenous variables for country i at year t that affect the quantity demanded of
international minutes. Again the initial observation for each country is at time t0, and the
last observation is at time t0+7. I include the following right-hand side variables:
INCOME; measured by real GDP per capita. If international telephone service is a
normal good I expect a positive relationship between income and the amount of minutes
per person. Importantly, whether international service is supplied by a monopolist or not
may affect output. Rate regulation would in theory make the monopolist price at Ramsey
**Draft**
11
levels and thus, increasing output and social welfare. In practice regulators face a large
problem of asymmetric information because they cannot observe nor elicit accurate
information on costs to the regulated firms. Given that rate regulation typically
incorporates an implicit rate of return for the regulated firm; inefficiency has not cost or
low cost to the regulated firm and productive efficiency declines (Sappington 1980).6
The maximum number of years of monopoly granted in my sample is ___ years.7
I assume that the behavior of monopolies may not be the same whether they hold
monopoly rights for, say two more years or 20 more years. Thus, I control for monopoly
using the remaining number of years of monopoly on international service
(INTLMONOP). I expect monopoly to be negatively correlated with output on
international telephony usage. To test the null hypothesis H01, I include a dummy
variable to indicate vertical separation between the owner of the local fixed network and
the international service provider. If there were several service providers I consider the
dominant firm in each case. I expect a negative relationship between vertical separation
and international telephony usage. Vertical separation increases the risk of opportunistic
behavior on either side, and this increases transaction costs (Williamson 1971, Klein,
Crawford, and Alchian 1978; Williamson 1985). Agreement between downstream and
upstream firms on access fees requires lengthy negotiations. Typically if no agreement is
reached within a reasonable timeframe, the regulator is charged with the duty of
determining the level of access fees to compensate the owner of the fixed local network
for using his property. It can hardly be emphasized the enormous information burden
6
Fir a good illustration of this and related problems see Berg and Tschirhart 1988: 505-511. For the
problem of productive inefficiency on monopolies see (Alchian and Kessel 1962).
7
See summary statistics in the appendix.
**Draft**
12
imposed on a regulator trying to estimate optimal Ramsey level access fees.8 Taking into
account that regulation is a politico-administrative process, rent-seeking costs would
increase losses due to productive inefficiency compounding the already negative
consequences of vertical separation (Tullock 1967). I also add to the right-hand side
variables the price of cellular telephone service.9 I have no prior expectations about the
sign of its coefficient as it will depend on whether cellular telephony is a complement
(positive sign) or a substitute (negative sign) of international telephony. Finally; I use a
year variable to control for technological change in the telecommunications industry. I
do not have enough degrees of freedom to include a dummy for each year so I use a
dummy for observations in the first half of the 1980s (1981-1985); a dummy for the
second half (1986-1990); and so on. The results appear in table 3. I report robust
standard errors that allow for within country (cluster) correlation assuming independence
between countries.
Using specification (3), the null hypothesis H01 is rejected with 99 percent
confidence. The purported benefits of baring the local fixed line operator from
international service do not exist. Vertical separation does not increase output on
international service. As expected, monopoly on international service is significantly
associated with a lower quantity of international telephone usage. The other variables
have non significant coefficients. Notice from specifications (1) and (2) that the results
for monopoly and vertical separation are fairly stable. From specification (3), we can
estimate the negative impact of vertical separation and monopoly on the quantity of
8
For a theoretical exposition of how to estimate Ramsey level access fees with full information see Laffont
and Tirole (2000: 80-83 and 97-105).
9
Explain how this is estimated…
**Draft**
13
international telephony. Countries that have vertically separated telephone firms exhibit
8.4 percent less international telephony usage per person than countries with integrated
firms;10 while one additional year of monopoly is associated with an 8.6 percent decline
on international telephony usage per person all other things constant.
Table 3: Effect of monopoly rights and vertical separation on outgoing international
minutes per person.
Variable
INTLMONOP
VERTSEP
Log[INCOME]
Log[CELLP2]
Constant
Observations
Countries
Country fixed effects
Year fixed effects
R-squared
Dependent variable = Log [INTLMIN]
(1)
(2)
(3)
-0.090
-0.090
(4.09)*
(4.07)*
-0.112
-0.077
(3.65)*
(2.56)**
0.306
1.094
0.303
(0.58)
(2.34)**
(0.57)
0.019
-0.106
0.019
(0.47)
(1.67)***
(0.46)
2.485
1.383
2.506
(2.02)**
(1.21)
(2.02)**
227
231
227
50
51
50
yes
yes
yes
yes
yes
yes
0.559
0.467
0.559
Panel data estimation with country fixed effects and year effects. Robust standard errors corrected for
heteroskedasticity and within country (cluster) correlation. t-statistics in parenthesis. * = 99 percent
confidence; ** = 95 percent confidence; *** = 90 percent confidence.
4. Consequences of vertical separation and monopoly on local fixed telephony
output and prices
The next step is to test the other two assertions made in the pre-privatization years:
10
The impact of vertical separation on international minutes per person (in percentage terms) is estimated
by: 100 (e-β-1). Where β is the estimated coefficient of VERTSEP. See Halvorsen and Palmquist (1980).
14
**Draft**
(1) monopoly helps advance universal residential local service (through cross
subsidization); and (2) introducing competition on basic services would lead to a break
down of the cross-subsidy scheme and thus an increase on residential local rates. I test
the following null hypotheses:
ƒ
H02: Monopoly on local fixed telephony is associated with more lines in service
than when competition is allowed
ƒ
H03: Monopoly on local fixed telephony is associated with lower residential
telephone rates than when competition is allowed.
To test H02 I use again panel data estimation with country fixed effects. I estimate the
following equation:
E [Log (FIXLINES )it | X it ]= X it β i + ε it ,
(2)
where, FIXLINESit is a vector of observations of the number of fixed lines in service per
person in country i at time t. As before, Xit is a matrix of exogenous variables for country
i at year t that affect the quantity of fixed lines in service. I include some of the same
right-hand side variables as before: INCOME; a dummy variable for monopoly in local
fixed telephony (LOCALMONOP); a dummy for vertical separation (VERTSEP); the
price of cellular telephony (CELLP2); and year dummy variables to control for
unobserved technological changes in the industry. I also add a variable to control for the
degree of fixed lines rationing. Typically, as the amount of subsidy to local residential
service increased, the state-owned monopolies were unable to serve the quantity
demanded at the subsidized price; thus a shortage (waiting lines) developed.11 At the
time of privatization many countries exhibited a large number of waiting lines.
11
Mention waiting time in some countries…..Ambrose? others?
**Draft**
Regulated prices by definition do not adjust instantly, so it took time to the new private
owners to eliminate this problem by adjusting prices to equate quantity supplied with
quantity demanded. I use the number of waiting lines per person to control for this. I
expect a negative relationship between this variable and the number of fixed lines per
person.
Table 4: Effect of monopoly rights and vertical separation on local fixed telephony.
Variable
LOCALMONOP
VERTSEP
Log[INCOME]
Log[WAITLINES]
Log[CELLP2]
Constant
Observations
Countries
Country fixed effects
Year fixed effects
R-squared
Dependent variable = Log [FIXLINES]
(1)
(2)
(3)
-0.050
-0.050
(4.05)*
(4.13)*
-0.304
-0.324
(7.07)*
(6.91)*
0.603
0.056
0.059
(1.60)
(0.15)
(0.16)
-0.106
-0.076
-0.070
(4.55)*
(2.57)**
(2.37)**
0.010
0.012
0.012
(1.37)
(2.20)**
(2.26)**
-3.667
-2.271
-2.183
(4.01)*
(2.39)**
(2.30)**
218
214
214
37
36
36
yes
yes
yes
yes
yes
yes
0.568
0.631
0.645
Panel data estimation with country fixed effects and year effects. Robust standard errors corrected for
heteroskedasticity and within country (cluster) correlation. t-statistics in parenthesis. * = 99 percent
confidence; ** = 95 percent confidence; *** = 90 percent confidence.
Table 4 presents the results for the quantity of fixed telephony. Using
specification (3), the null hypothesis H02 is rejected with 99 percent confidence level.
15
**Draft**
16
Monopoly does not promote universal service in local telephony.12 Countries that have
the local fixed network providers separated from the international service provider
exhibit significantly less fixed lines in service. Therefore, vertical separation not only
harms the downstream international service provision but also the provision of the
upstream local fixed telephony service. As Crandall and Sidak (2004: 410) conclude in
the context of the U.S. experience: “Mandatory structural separation is unnecessary
because the putative benefits that it would produce are, in fact nonexistent.”
As expected, countries that have a higher number of waiting lines per person
exhibit significantly lower number of fixed lines per person. Consistent with previous
findings (CITE..) the coefficient of the price of cellular telephony indicates that fixed
lines and mobile telephony are close substitutes. The results for monopoly and vertical
separation appear fairly robust to different specifications reported in table 4. Changing
the monopoly variable to monopoly in all basic services (i.e., local, long distance, and
international) does not alter the results. Monopoly does not advance universal service in
local fixed telephony it retards it. From specification (3), we can estimate the negative
impact of monopoly and vertical separation on the quantity of fixed lines in service per
person. Countries that have vertically separated telephone firms exhibit 27.7 percent less
lines in service per person than countries with integrated firms;13 while one additional
year of monopoly is associated with a five percent decline on the number of fixed lines
per person all other things constant.
Finally I test H03 and again I use a regression equation similar to (2) but using the
12
Wallsten (2004) found similar results; but for the most part lacked significance.
The impact of vertical separation on fixed lines in service per person (in percentage terms) is estimated
by: 100 (e-β-1). See Halvorsen and Palmquist (1980).
13
**Draft**
17
natural logarithm of the price of local residential telephony as the dependent variable.
The right-hand side variables are similar as before but I add the degree of digitalization of
the local network (DIGITAL). As digital technology is deployed the quality of the
residential local services increases. For example, the clarity of voice communications
may be enhanced, new services such as voice mail and call waiting may be offered as
part of a package for local telephony which may affect residential local prices. Table 5
presents the results using two measures for the price of local residential telephony. Using
specification (2) I reject with 95 confidence level H03. Monopoly in local telephony is
associated with higher prices not lower prices of local residential telephony. The preprivatization arguments in favor of monopoly to avoid a sharp increase on the prices of
the subsidized residential telephone service were unwarranted. Analysts underestimated
the power of private property and competition to enable gains in productive efficiency
that would more than compensate any increase on prices above marginal cost in the postprivatization years.
Vertical separation, once again harms the expansion of basic telephony by
producing an increase on prices holding all else constant. Again this can be the results of
inefficiencies arising from separating firms that invest in assets that are interrelated or
specific (CITE…) Other variables came with the expected sign. It is important to notice
that the results are fairly robust to alternative specifications and alternative indicators of
residential prices. Changing the monopoly variable to indicate monopoly in all basic
services does not alter the conclusions. Using specification (2) each additional year of
monopoly increases the price of residential local service by 10.2 percent, and vertical
separation increases residential prices by 12.3 percent.
**Draft**
18
Table 5: Effect of monopoly rights on the price of local residential fixed telephony.
Variable
LOCALMONOP
VERTSEP
Log[INCOME]
Log[DIGITAL]
Log[CELLP2]
Constant
Observations
Countries
Country fixed effects
Year fixed effects
R-squared
Dependent variable
Log[LOCALRES2]
Log[LOCALRES1]
(1)
(2)
(3)
(4)
0.102
0.102
0.113
0.113
(2.05)**
(2.04)**
(2.22)** (2.21)**
0.116
0.143
(1.88)***
(2.19)**
1.139
1.142
1.337
1.341
(0.95)
(0.95)
(1.04)
(1.04)
0.369
0.370
0.356
0.358
(2.32)**
(2.31)**
(2.00)** (1.99)**
0.087
0.087
0.084
0.084
(1.23)
(1.23)
(1.15)
(1.15)
0.453
0.422
-0.220
-0.259
(0.15)
(0.14)
(0.07)
(0.08)
246
246
246
246
55
55
55
55
yes
yes
yes
yes
yes
yes
yes
yes
0.283
0.283
0.274
0.274
Panel data estimation with country fixed effects and year effects. Robust standard errors corrected for
heteroskedasticity and within country (cluster) correlation. t-statistics in parenthesis. * = 99 percent
confidence; ** = 95 percent confidence; *** = 90 percent confidence.
5. Conclusions
Two myths have been dispelled analyzing the post-privatization experience of
basic telephony around the world. The first myth is that monopoly is needed to maintain
low (subsidized) prices on local residential telephony. The second myth is that vertical
separation of the local fixed network from the provision of international telephony (or
long distance) is needed to prevent a dominant vertically integrated firm from excluding
rivals in the downstream services (i.e., international telephony). Thus, vertical (or
structural) separation promised more competition and increased output on this market.
The results of this article show that monopoly harms upstream and downstream telephone
**Draft**
19
services, and increases the price of residential telephony a traditionally subsidized
service. Vertical separation also harms the expansion of upstream and downstream
telephone services. Therefore, monopolizing basic telephony services does not advance
universal service because it increases prices and reduces the number of fixed lines in
service. Vertical separation has the same effect; it reduces the expansion of upstream and
downstream basic telephone services. Both policy options reduce social welfare and
harm those consumers that were precisely designed to help: the downstream users of
international telephony and the upstream users of residential local fixed telephony.
**Draft**
20
REFERENCES
Alchian, Armen A. 1965. Some Economics of Property Rights. Il Politico. 30 (4): 816829.
Alchian, Armen A. 1969. Corporate Management and Property Rights. in Economic
Policy and the Regulation of Corporate Securities, Manne, Henry, ed.
(Washington, DC: The American Enterprise Institute).
Alchian, Armen A., and Reuben A. Kessel. 1962. Competition, Monopoly and the
Pursuit of Money. in Aspects of Labor Economics. National Bureau of Economic
Research (Princeton, NJ: Princeton University Press).
Aviram, Amitai. 2003. “Regulation by Networks.” Bringham Young Law Review. Vol.
2003 (4): 1179-1235.
Berg, Sanford V. and John Tschirhart. 1988. Natural Monopoly Regulation. Principles
and Practice. (Cambridge, UK: Cambridge University Press)
Bös, Dieter. 1993. “Privatization in Europe: A Comparison of Approaches.” Oxford
Review of Economic Policy, Vol. 9 (1): 95-111.
Coase, R. H. 1937. “The Nature of the Firm.” Economica. Vol. 4 (16): 386-405.
Comanor, William S., and Harvey Leibenstein, 1969. Allocative Efficiency, XEfficiency and the Measurement of Welfare Losses. Economica. 36 (143): 304309.
Crandall, Robert W. 1988. Surprise from Telephone Deregulation and the AT&T
Divestiture. The American Economic Review. Vol. 78 (2): 323-327.
Crandall, Robert W. 1989. “The Role of the U.S. Local Operating Companies.” In
Crandall, Robert W. and Kenneth Flamm, eds., Changing the Rules.
(Washington, DC: The Brookings Institution).
Crandall, Robert W. and J. Gregory Sidak. 2002. “Is structural Separation of Incumbent
Local Exchange Carriers Necessary for Competition?” Yale Journal on
Regulation. Vol. 19: 335-411.
Cremer, Jacques; Patrick Rey and Jean Tirole. 2000. “Connectivity in the Commercial
Internet.” The Journal of Industrial Economics. Vol. 48(4): 433-472.
Cronin, Francis J.; Elisabeth Colleran; Michael Miller and Richard Raczkowski. 1997.
“Local Exchange Competition, Rate Restructuring, and Universal Service.
Telecommunications Policy. Vol. 21 (3): 251-264.
Faulhaber, Gerald R. 1975. Cross-Subsidization: Pricing in Public Enterprises. The
**Draft**
21
American Economic Review. Vol. 65 (5): 966-977.
Furubotn, Eirik G., and Svetozar Pejovich, 1972. Property Rights and Economic Theory:
A survey of Recent Literature. Journal of Economic Literature. 10 (4): 11371162.
Grossman, Sanford J. and Oliver D. Hart. 1986. The Costs and Benefits of Ownership: A
Theory of Vertical and Lateral Integration. The Journal of Political Economy.
Vol. 94 (4): 691-719.
Halvorsen, Robert, and Raymond Palmquist. 1980. The Interpretation of Dummy
Variables in Semilogarithmic Equations. The American Economic Review. Vol.
70 (3): 474-475.
Hausman, Jerry, Timothy Tardiff and Alexander Belinfante. 1993. “The Effects of the
Breakup of AT&T on Telephone Penetration in the United States.” The American
Economic Review. Vol. 83 (2): 178-184.
Klein, Benjamin; Robert G. Crawford and Armen A. Alchian. 1978. “Vertical
Integration, Appropriable Rents, and the Competitive Contracting Process.
Journal of Law and Economics. Vol. 21(2): 297-326.
Laffont, Jean-Jacques and Jean Tirole. 2000. Competition in Telecommunications.
(Cambridge, MA: The MIT Press)
Leibenstein, Harvey, 1966. Allocative Efficiency vs. “X-Efficiency.” The American
Economic Review. 56 (3): 392-415.
Nambu, Tsuruhiko, Kazuyuki Suzuki and Tetsushi Honda. 1989. “Deregulation in
Japan.” In Crandall, Robert W. and Kenneth Flamm (eds.). Changing the Rules.
(Washington, DC: The Brookings Institution).
Palmer, Karen. 1992. “A Test for Cross Subsidies in Local Telephone Rates: Do Business
Customers Subsidize Residential Customers?” The Rand Journal of Economics.
Vol 23 (3): 415-431.
Pilcher, Joseph E. 1994. “The Point of view of a Global Operator: Cable & Wireless.”
In Wellenius, Bjorn and Peter A. Stern, eds., Implementing Reforms in the
Telecommunications Sector. Lessons form Experience. Washington, DC: The
World Bank.
Salop, Steven C. and David T. Scheffman. 1983. Raising Rivals’ Costs. The American
Economic Review. Vol. 73 (2): 267-271.
Sappington, David. 1980. Strategic Firm Behavior Under a Dynamic Regulatory
Adjustment Process. The Bell Journal of Economics. Vol. 11 (1) 360-372.
**Draft**
22
Tullock, Gordon. 1967. The Welfare Costs of Tariffs, Monopolies, and Theft. Western
Economic Journal. Vol. 5 (June): 224-232.
Wallsten, Scott J. 2004. “Privatizing Monopolies in Developing Countries: The Real
Effect of Exclusivity Periods in Telecommunications.” Journal of Regulatory
Economics. Vol. 26 (3): 303-320.
Waterschoot, Paul. 1994. “Regulation and Competition Policy. In Wellenius, Bjorn and
Peter A. Stern, eds., Implementing Reforms in the Telecommunications Sector.
Lessons form Experience. Washington, DC: The World Bank.
Williamson, Oliver E. 1963. Managerial Discretion and Business Behavior. The
American Economic Review. 53 (5): 1032-1057.
Williamson, Oliver E. 1971. The Vertical Integration of Production: Market Failure
Considerations. The American Economic Review. Vol. 61 (2): 112-123.
Williamson, Oliver E. 1985. The Economics Institutions of Capitalism. (New York, NY:
The Free Press).
Whinston, Michael D. 1990. Tying, Foreclosure, and Exclusion. The American
Economic Review. Vol. 80 (4): 837-859.