sibling rivalry: the emergence of competition among the baby bells
TRANSCRIPT
MANAGERIAL AND DECISION ECONOMICS, VOL. 16,479-492 (1995)
Sibling Rivalry: The Emergence of Competition
Among the Baby Bells William F. Shughart I1
Uniwrsity of Mississippi, MS, USA
The Modification of Final Judgment (MFJ) prohibits the seven regional Bell operating companies (RBOCs) spun off by AT&T from competing with one another in various markets ‘adjacent’ to the local telephone exchange. I have examined the competitiveness of three of these a4acent markets (cellular telephone, paging services, and yellow pages publishing) since divestiture in order to assess the likelihood of collusion among the RBOCs. On the basis of a decade of post-divestiture experience, I find the prospect of collusion to be remote and, hence, the probability of anticompetitive effects from lifting the MFJ to be vanishingly small.
INTRODUCTION
I have been asked to assess the likelihood that the Regional Bell Operating Companies (RBOCs), if freed from the line-of-business re- strictions of the Modification of Final Judgment (MFJ), could obtain monopoly power in markets adjacent to the local telephone exchange. I con- clude that there is no likelihood of such monopol- ization either through independent action by the RBOCs or collusion among them. On the basis of a decade of post-divestiture experience, it is ap- parent that no one RBOC is large enough to monopolize either those markets in which it is a buyer (central office switching equipment, for in- stance) or those in which it would be a seller (such as interexchange services) if the MFJ’s line- of-business restrictions were lifted. Furthermore, substantial regulatory safeguards have become available since 1984 to prevent any unilateral action that might result in a lessening of competi- tion.
While the history of the Bell System may have made concerns about collusion between the RBOCs understandable at the time the MFJ was issued, no evidence of collusion has materialized in the ten years since. The seven RBOCs have consistently acted independently since the breakup of AT&T, competing vigorously with one another at every opportunity. I find the prospect of collusion to be remote and, hence,
CCC 0143-6570/95/040479-14 0 1995 by John Wiley & Sons, Ltd.
the probability of anticompetitive effects from lifting the MFJ to be vanishingly small. Thus, I strongly support removing the line-of-business re- strictions imposed by Judge Greene.
THE ANTICOMPETITIVE THEORIES UNDERLYING THE MFJ
Since 1984, when the MFJ went into effect, its line-of-business restrictions have prohibited the RBOCs from offering interexchange telecommu- nications services, from manufacturing telecom- munications equipment, and from distributing telecommunications equipment other than so- called customer premise equipment (CPE). Addi- tional MFJ restrictions imposed on RBOC parti- cipation in information services and non-telecom- munications businesses have been lifted already.
The fears that led to the line-of-business re- strictions can be classified under two broad head- ings. These are cross-subsidization and foreclo- sure.
Cross-subsidization involves the REBOW strate- gic use of their control over local exchange facili- ties to finance predatory behavior in adjacent markets or to increase returns from regulated businesses. This could be accomplished, first, by shifting disproportionate shares of the common costs of supplying related services, say local and
480 WILLIAM F. SHUGHART II
long-distance telephone calls, to the local ex- change’s rate base. Where local telephone rates are set by regulators based on the costs of provid- ing service, some of the costs of supplying long- distance service would in this hypothetical exam- ple be passed on to consumers in the form of higher rates for local telephone service. Because it would thereby earn higher returns from regu- lated operations, a RBOC could profitably charge below-cost long-distance rates. Such cross-sub- sidization may allow the RBOC to drive out exist- ing rivals and to strategically deter the entry of new ones, enabling it eventually to earn supra- competitive returns in the long-distance market.
The MFJ assumes that the RBOCs may also be able to subsidize competitive businesses and cir- cumvent rate regulation through a form of self- dealing. Suppose the RBOCs are permitted to manufacture telephone switches, for example. An incentive would then exist to set the intracom- pany (transfer) prices of switches or other inter- nally supplied inputs above their marginal pro- duction cost. These above-marginal-cost input prices would be added, at least in part, to the local telephone exchange’s rate base, and conse- quently shifted forward to local rate payers. Self- dealing of this sort was in fact one of the major allegations of anticompetitive behavior levied against the old vertically integrated Bell System?
The second fear underlying the MFJ’s line-of- business restrictions, namely foreclosure, assumes that the RBOCs could deny rivals in adjacent markets access to the local exchange, or grant access at discriminatory prices. Through such dis- crimination, a RBOC could theoretically raise the cost of competitor’s products and services or drive competitors from a market entirely, enabling the RBOC to increase its own profits in that market.
None of these concerns necessarily depends upon collective action by the RBOCs. Neverthe- less, the possibility of collusion among the RBOCs was thought to add significantly to the competi- tive risk of allowing RBOCs into adjacent mar- kets. Concerns about RBOC collusion extended particularly to manufacturing markets. The RBOCs, it has been said, might ‘act in concert with respect to manufacturing and purchasing’ and ‘enter into explicit or implicit agreements with each other regarding specifications or inter- connection req~irements’.~ Such cooperation, the theory goes, would allow the RBOCs to drive other competitors out of adjacent markets and
then to divide those markets among themselves, enabling the realization of supracompetitive pro- fits? Similar strategies might be employed in in- terexchange markets, where FtBOCs could theo- retically combine to raise the costs of rival carri- ers and drive them from the market.
In theory, collusion might also facilitate cross- subsidization. Absent collusion, a single RBOC‘s effort to raise the prices of inputs to exchange service above marginal cost would likely cause other producers of the inputs to undercut that RBOC‘s prices. Likewise, regulators could readily detect a RBOC‘s padded costs through compar- isons with benchmark cost information obtained from the other six RBOCs and non-RBOC local exchange carriers. If, however, the m0Cs agreed not to offer inputs at competitive prices or to report similarly inflated costs, it would be some- what easier to conceal and sustain cross-subsidies.
OBSTACLES TO ANTICOMPETITIVE CONDUCT
Concerns about cross-subsidy, foreclosure, and collusion may have seemed well founded at a time when the Bell System dominated nearly all telecommunications markets and overwhelmed regulators with its sheer size. Today, however, the fears that prompted the line-of-business restric- tions are comprehensively addressed by regula- tion and market forces.
Since 1984, incentive-based rate regulation has been adopted at the federal level and has become much more widely used at the state level. So-called price-cap regulation, for instance, limits allowable rate increases to a specifled percentage that only partially offsets increases in the overall rate of inflation. Such regulatory pricing formulas create stronger incentives to economize on the costs of providing service than older forms of rate-of-re- turn regulation by limiting the extent to which cost increases can be passed on to consumers in the form of higher prices. Structural and account- ing separation policies have also been adopted that sharply reduce opportunities to evade rate regulation through cross-subsidization. In addi- tion, non-RBOC suppliers of telephone equip- ment and other inputs provide regulators with competitive benchmarks for determining the rea- sonableness of the RBOCs’ internal transfer prices.
THE EMERGENCE OF COMPETITION AMONG THE BABY BELLS 481
Equal access and interconnection rules address fears that the RBOCs might deny competitors access to the local exchange, provide them with inferior access, or charge them discriminatory prices. The Federal Communications Commission (FCC), for instance, has established technical and performance standards that CPE must meet in order to be connected to the local network. If a rival's equipment meets these standards, the RBOC cannot deny connection. The FCC also requires that all telcos offer local service unbun- dled from CPE. This provision precludes the RBOCs from either requiring customers to pur- chase RBOC-manufactured equipment in order to receive local service, or charging customers more for local service if the equipment is not manufactured by an affiliated supplier. Disclosure and structural separation requirements protect against the possibility that the RBOCs will afford their own competitive business operations prefer- ential access to network or customer information. These rules make the possibility of foreclosure remote.
Finally, opportunities to cross-subsidize unregu- lated operations, or to deny competitors equal access, promise to be constrained even more tightly by actual and potential competition in the supply of local telephone service. The develop ment of competitive access providers, wireless telephone technology, and other methods of by- passing the local exchange, as well as the major interexchange carriers' steps toward building their own local networks, cast doubt on the whole notion that a local exchange 'bottleneck' exists in the first place.
IMPEDIMENTS TO COLLUSION AND MARKET EXPERIENCE
By erecting formidable barriers to cross-subsidi- zation and foreclosure, the regulatory and market developments discussed above indirectly address many of the concerns about collusion among the RBOCs. In addition, anticompetitive collusion is directly forbidden by federal and state antitrust laws, violation of which carries substantial penal- ties. Civil and criminal penalties for violations of the Sherman Act, for example, include treble damages: corporate h e s of up to $10 million (along with personal fines of as much as
$350000),6 and mandatory minimum prison sen- tences?
In addition, theories that the RBOCs might collude to effectively revive the Bell System and its supposed anticompetitive practices overlook practical realities. The RBOCs are no longer part of a single entity; they are seven independent and competitive companies. This simple fact renders successful collusion highly improbable.
Firms collude in order to increase their joint profits. This same profit motive provides powerful incentives for the individual members of a price- fixing conspiracy to cheat on their agreement: One cartel member can increase its own profits at the expense of the others by secretly expanding output and cutting price. Hence, for collusion to be successful, the conspirators might be able to detect and to punish the inevitable cheating on their agreement; they must also act collectively to deter the entry of new rivals who wish to share in the cartel's profits!
Successful collusion also requires that the mem- bers of a price-fixing conspiracy agree on a method of dividing the market among themselves. Assign- ing geographic territories is the most likely method of market division because extra-terri- torial sales then become prim facie evidence of cheating (Stigler, 1964). Indeed, the MFJ Court has speculated that the RBOCs would engage in just this sort of territorial division if allowed into telecommunications manufacturing markets?
Finally, it is probable that successful collusion would necessarily involve several RBOCs, not just two of them. The RBOCs collectively switch 85% of all interexchange calls and control 77% of all customer access lines. But the largest RBOC, Bell Atlantic, switches only 16% of all interstate calls and controls only 12.5% of all access lines; South- western Bell accounts for less than 9% of inter- state calls and access lines. (See Tables 1 and 2). Individual shares of this magnitude are insuffi- cient for effective predation in adjacent markets;" even a combination of two RBOCs likely would be inadequate. Only by combining several RBOCs' shares of exchange access or input purchases, and several RBOCs' war chests for subsidizing entry deterrence, could the RBOCs create a genuine anticompetitive threat.
Evidence of Rivalry It obviously is impossible to prove (or disprove)
482 WILLIAM F. SHUGHART I1
Table 1. RBOC Shares of Tier I Interstate Switched Access Minutes, 1991
Company
Bell Atlantic BellSouth "Ex Ameritech US West Southwestern Bell Pacific Telesis Total
Shire (46)
15.9 15.2 13.3 11.4 11.3 8.9 8.1
84.1
Some: US Federal Communications Commission (1993, pp. 212-13)
that the RBOCs could not successfully engage in anticompetitive collusion if the remaining MFJ restrictions were removed. Nevertheless, we now have up to a decade of experience with RBOC participation in adjacent markets such as cellular services, paging services, and publication of yel- low page directories. Concerns about RBOC col- lusion can be tested directly against this experi- ence.
If the RBOCs are able to overcome the formidable obstacles to collusion in adjacent mar- kets, one would expect them to agree to remain in their home territories, or at least to eschew inter-RBOC competition. Evidence of RBOC ver- sus RBOC competition does not absolutely rule out the possibility of collusion, but it does make the existence of such collusion significantly less likely. Inter-RBOC rivalry also complicates the negotiation and enforcement of future collusive agreements, making fears of anticompetitive cooperation that much less plausible.
Table 2. RBOC Shares of Local Exchange Callers (LEC) Access Lines, 1992
Company Share (%)
Bell Atlantic BellSOUth Ameritech "Ex Pacific Telesis us west Southwestern BeU Total
12.5 125 11.7 10.8 10.0 9.2 8.7
75.4
sourcC: US Telephone Association 1993, pp. 21-7.
In cellular telephone services, there is substan- tial RBOC versus RBOC competition in a num- ber of markets within each RBOC region. (See Table 3). As of 1992, for instance, Southwestern Bell competed against Ameritech in Illinois, Indi- ana, and Missouri; against Bell Atlantic in Mary- land, Virginia, and the District of Columbia; and against " E X in Massachusetts. To give an- other example, BellSouth faced competition within its home region from Bell Atlantic and Pacific Telesis, but competed against Ameritech and Pacific Telesis in their home regions.
Similar competition appears to exist in paging and yellow pages publishing markets. In fact, the incidence of RBOC versus RBOC competition is even greater in paging services than in cellular communications. As shown in Table 4, the RBOCs compete with one another in all but two of the nation's 105 largest metropolitan areas. Similarly, as shown in Table 5, six of the seven RBOCs face substantial competition in yellow pages publishing from at least one other RBOC that has entered their home-region market. This competition is hard evidence that the RBOCs will not collude along regional lines.
The RBOCs are also positioning themselves to compete against one another in the delivery of voice, video and data transmission services at the local level. Alliances with cable television compa- nies provide the RBOCs with the ability to supply these services outside their home regions. In 1993, all the RBOCs except Ameritech announced plans to enter into alliances with cable TV companies outside their home regions." While proposed al- liances have not always been consummated, such developments demonstrate that competition between RBOCs will extend even to the RBOCs' local telephony markets.'* This evidence that RBOCs do not collude with
one another to impede competition is reinforced by shifts in the RBOCs' individual shares of the cellular market. If the RBOCs have been collud- ing, one would expect each individual RBOC's share of total RBOC cellular subscribership to remain relatively stable over time. The facts do not show such share stability, however, as evi- denced by the information displayed in Fig. l. Between 1988 and 1992, Pacific Telesis' share of total RBOC cellular telephone subscribers fell from 22% to 13%. At the other extreme, South- western Bell's share rose from 19% to 25%. This
THE EMERGENCE OF COMPETITION AMONG THE BABY BELLS 483
Table 3. RBOC versus RBOC Competition in Cellular Markets, 1992
Reaon State
Ameritech WI WI WI WI WI WI MI MI OH OH OH OH OH OH OH IL IL IL IL IL IL IL
Bell Atlantic
Bell South
" E X
Pacific Telesis
DC MD VA
sc GA GA GA GA
MA MA MA MA MA RI RI
CA CA
Southwestern MO Bell MO
MO MO
Mprket location
Milwaukee Madison Racine Janesville Kenosha Sheboygan Detroit F h t Cincinnati Columbus Dayton Hamilton Springfield OH 4 OH 8
Springfield Champaign Decatur Bloomington IL 6 Gary-Hammond
Washington Baltimore
Chicago
VA 10-B1
Columbia Atlanta Athens GA 3 GA 4-B1
Boston New Bedford Pittsfield Worcester M A 2 Providence R I l
Los Angeles San Diego
St. Louis MO 8 MO 18 MO 19
RBOC oompetitor
liccnscc
BellSouth BellSouth BellSouth BellSouth BellSouth BellSouth PacTel PacTel PacTel PacTel PacTel PacTel PacTel PacTel PacTel Southwestern Bell Southwestern Bell Southwestern Bell Southwestern Bell Southwestern Bell Southwestern Bell Southwestem Bell
Southwestern Bell Southwestern Bell Southwestern Bell
Bell Atlantic PacTel PacTel PacTel PacTel
Southwestern Bell Bell Atlantic Bell Atlantic Southwestern Bell Southwestern Bell Bell Atlantic Bell Atlantic
BellSouth us west
Ameritech Ameritech Ameritech Ameritech
Source: Cellular Telecommunications Industry Association (1992)
Tab
le 4
. R
BOC
ver
sus
RBO
C C
ompe
titio
n in
Pag
ing
Serv
ices
, 105
Lar
gest
Met
ropo
litan
Sta
tistic
al Ar
eas, 1
993
No. 1 2 3 4 5 6 7 8 9 10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Sta
te
New
Yor
k C
alifo
rnia
Ill
inoi
s Pe
nnsy
lvan
ia
Mic
higa
n M
assa
chus
etts
C
alifo
rnia
Dc
Texa
s Te
xas
Miss
ouri
Flor
ida
Penn
sylv
ania
M
aryl
and
Min
neso
ta
Ohio
Geo
rgia
C
alifo
rnia
C
olor
ado
Was
hing
ton
Wisc
onsin
Fl
orid
a Ohio
Miss
ouri
New
Yor
k A
rizon
a C
alifo
rnia
In
dian
a Lo
uisi
ana
Ore
gon
Ohio
Texa
s N
ew Y
ork
Cal
iforn
ia
Tenn
esse
e K
entu
cky
Rho
de Is
land
U
tah
Ohio
&M
dCU
t
City
New
Yor
k L
os A
ngel
es
Chi
cago
Ph
ilade
lphi
a D
etro
it B
osto
n Sa
n Fr
anci
sco
Was
hing
ton
Dal
las
Hou
ston
St
Lou
is
Miami
Pitts
burg
h B
altim
ore
Min
neap
olis
C
leve
land
A
tlant
a Sa
n D
iego
D
enve
r Se
attle
M
ilwau
kee
Tam
pa
Cin
cinn
ati
Kansas C
ity
Buf
falo
Ph
oeni
x Sa
n Jo
se
Indi
anap
olis
N
ew O
rlean
s Po
rtlan
d C
olum
bus
Har
tford
Sa
n A
nton
io
Roc
hest
er
Sacr
amen
to
Mem
phis
Lo
uisv
ille
Prov
iden
ce
Salt
Lake
City
D
ayto
n
RB
OC
regi
on
"E
x
PacT
el
Am
erite
ch
Bel
l Atla
ntic
A
mer
itech
"E
x
PacT
el
Bell
Atla
ntic
sw
Bell
SW B
ell
sw Bell
Bel
lSou
th
Bel
l Atla
ntic
B
ell A
tlant
ic
us W
est
Am
erite
ch
Bel
lsou
th
PacT
el
US
Wes
t us
wes
t A
mer
itech
B
ellS
outh
A
mer
itech
sw
Bell
"E
x
SW B
ell
PacT
el
Am
erite
ch
Bel
lSou
th
PacT
el
Am
erite
ch
"EX
sw Bel
l "EX
PacT
el
Bel
lSou
th
Bel
lSou
th
"EX
US W
est
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
Bel
l Atla
ntic
A
mer
itech
B
ellS
outh
B
ellS
outh
B
ell A
tlant
ic
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
Bel
lSou
th
Bel
l Atla
ntic
B
ell A
tlant
ic
Am
erite
ch
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
Bel
lsou
th
Am
erite
ch
Bel
lSou
th
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
Bel
lsou
th
Bel
lSou
th
Bel
lSou
th
PacT
el
Bel
lSou
th
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
RB
OC
pag
ing service
prov
ider
s
"EX
PacT
el
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
sw Bel
l Pa
cTel
Be
llsou
th
PacT
el
PacT
el
Bel
lsou
th
PacT
el
Bel
l Sou
th
Bel
lSou
th
Bel
lsou
th
Bel
lSou
th
PacT
el
PacT
el
PacT
el
PacT
el
Bel
lSou
th
PacT
el
Bel
lSou
th
PacT
el
"E
x
PacT
el
PacT
el
Bel
lSou
th
PacT
el
PacT
el
Bel
lSou
th
PacT
el
PacT
el
"E
x
sw Bel
l Pa
cTel
B
ellS
outh
"EX
PacT
el
Bellsouth
PacT
el
sw Bel
l Pa
cTel
Pa
cTel
Pa
cTel
sw Bell
PacT
el
sw Bel
l sw
Bell
PacT
el
sw Bel
l Pa
cTel
Pa
cTel
Pa
cTel
Pa
cTel
sw
Bell
sw Be
ll sw
Bell
us W
est
PacT
el
sw Bel
l Pa
cTel
sw
Bell
PacT
el
sw Be
ll SW
Bel
l Pa
cTel
sw
Bell
us W
est
PacT
el
sw Be
ll sw
Bell
PacT
el
PacT
el
sw Be
ll US W
est
PacT
el
8
P
sw Be
ll
sw Be
ll sw
Bell
sw Bel
l
us W
est
sw Bel
l
sw Bell
sw Bel
l sw
Bell
sw Bel
l 4 F
us w
est
iz US W
est
.o us
Wes
t
sw Bell
sw Bell
+I
z i us
wes
t
sw Bell
sw Bell
sw Be
ll
sw Bel
l
sw Bel
l
Tab
le 4
. R
BOC
ver
sus
RBO
C C
ompe
titio
n in
Pag
ing
Serv
ices
, 105
Lar
gest
Met
ropo
litan
Sta
tistic
al Ar
eas, 1
993
(Con
t.)
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Ala
bam
a C
onne
ctic
ut
Virg
inia
N
ew Y
ork
Okl
ahom
a Te
nnes
see
Nor
th C
arol
ina
Ohio
Haw
aii
Florida
Ohi
o N
ew Y
ork
Indi
ana
Mas
sach
uset
ts
Penn
sylv
ania
O
klah
oma
Penn
sylv
ania
V
irgin
ia
Florida
Nor
th C
arol
ina
New
Jer
sey
Mas
sach
uset
ts
Neb
rask
a Ohio
Sout
h C
arol
ina
Del
awar
e N
ew J
erse
y N
orth
Car
olin
a Florida
Cal
iforn
ia
Cal
ifom
aia
Texa
s M
assa
chus
etts
A
rizon
a
CoM
&kU
t
Mic
higa
n
Mic
higa
n
Birm
ingh
am
Brid
gepo
rt N
orfo
lk
Alb
any
Okl
ahom
a City
N
ashv
ille
Gre
ensb
oro
Tole
do
New
Hav
en
Hon
olul
u Ja
ckso
nvill
e A
kron
*a- Ga
ry-H
ammo
nd
Wor
cest
er
NE P
enns
ylva
nia
Tul
sa
Alle
ntow
n R
ichm
ond
Orla
ndo
Cha
rlotte
N
ew B
runs
wic
k Sp
ringf
ield
G
rand
Rap
ids
Om
aha
Gre
envi
lle
Flin
t W
ilmin
gton
Lo
ng B
ranc
h R
alei
gh
Wes
t Pal
m B
each
h
ar
d
Fres
no
AU
Stin
N
ew B
edfo
rd
Tucs
on
You
ngst
own
Bel
lSou
th
"E
x
Bell
Atla
ntic
"-EX
sw Bel
l B
ells
outh
B
ellS
outh
A
mer
itech
"E
x
Bel
lsou
th
Am
erite
ch
"E
x
Am
erite
ch
"E
x
Bell A
tlant
ic
SW B
ell
Bell
Atla
ntic
Bell A
tlant
ic
Bel
lsou
th
Bel
lsou
th
"E
x
"E
x
Am
erite
ch
us w
est
Am
erite
ch
Bel
lsou
th
Am
erite
ch
Bell
Atla
ntic
"E
x
Bel
lsou
th
Bel
lSou
th
PacT
el
PacT
el
sw Be
ll "E
x
sw Bell
Bel
lSou
th
Bel
lSou
th
Bell A
tlant
ic
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
Bel
l Sou
th
Am
erite
ch
Bel
lSou
th
Bel
lSou
th
Bel
lSou
th
Am
erite
ch
Bel
lsou
th
Am
erite
ch
Bel
lsou
th
Bel
l Atla
ntic
B
ellS
outh
B
ell A
tlant
ic
Bel
l Atla
ntic
B
ellS
outh
B
ellS
outh
B
ell A
tlant
ic
Bel
lSou
th
Am
erite
ch
Bel
lSou
th
Am
erite
ch
Bel
lsou
th
Am
erite
ch
Bell A
tlant
ic
Bell A
tlant
ic
Bel
lsou
th
Bel
lSou
th
Bel
lsou
th
Bel
lSou
th
Bel
lsou
th
Bel
lsou
th
Bel
lsou
th
PacT
el
sw Be
ll B
ell S
outh
Pa
cTel
Pa
cTel
Pa
cTel
Pa
cTel
B
ells
outh
Pa
cTel
PacT
el
Bellsouth
"E
x
Bel
lSou
th
sw Be
ll B
ellS
outh
Pa
cTel
B
ellS
outh
B
ellS
outh
Pa
cTel
Pa
cTel
sw
Bell
PacT
el
Bel
lsou
th
PacT
el
PacT
el
PacT
el
Bel
lSou
th
Bel
lsou
th
sw Bel
l Pa
cTel
Pa
cTel
Pa
cTel
Pa
cTel
Pa
cTel
sw
Bell
PacT
el
PacT
el
SW B
ell
SW B
ell
sw Bel
l
sw Bell
sw Bel
l Pa
cTel
Pa
cTel
Pa
cTel
sw Be
ll Pa
cTel
Pa
cTel
SW
Bel
l
sw Bell
sw Be
ll us
Wes
t
PacT
el
sw Bel
l
sw Be
ll sw
Bell
sw Be
ll sw
Bell
sw Bel
l
sw Be
ll
sw Bel
l sw
Bell
sw Bel
l
E 4
W E
us W
est
P
W
o\
Tab
le 4
. pR
BO
C~e
rsus
GO
C C
om
pet
itio
nin
P& Se
rvic
es, 1
05 L
arge
st M
etro
polit
an S
tatis
tical
Are
as, 1993 (Cont.)
78
79
80
81
82
83
a4
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
Michigan
Lam
ing
Am
erite
ch
Am
erite
ch
Bel
lSou
th
PatT
el
sw Bel
l Te
nnes
see
Kn
de
B
ellS
outh
B
ellS
outh
Pa
cTel
sw Bell
Loui
sian
a B
aton
Rou
ge
Bel
lSou
th
BellSouth
PacT
el
sw Bel
l Te
xas
El P
as0
sw Be
ll
Bel
lSou
th
PacT
el
sw Bell
Was
hing
ton
Taco
ma
us W
est
Bel
lSou
th
PacT
el
us W
est
Ala
bam
a M
obile
B
ellS
outh
B
ellS
outh
Pa
cTel
Pe
nnsy
lvan
ia
Har
risbu
rg
Bel
l Atla
ntic
B
ell A
tlant
ic
Bel
lSou
th
Tenn
esse
e Jo
hnso
n C
ity
Bel
lSou
th
Bel
lSou
th
PacT
el
New
Mexico
Alb
uque
rque
us
wes
t B
ellS
outh
Pa
cTel
sw Bell
Ohi
o C
anto
n A
mer
itech
A
mer
itech
B
ellS
outh
Te
nnes
see
Cha
ttano
oga
BellSouth
Bel
lSou
th
Kan
sas
Wic
hita
sw
Bell
Bel
lSou
th
PacT
el
sw Be
ll
Sout
h C
arol
ina
Cha
rlest
on
Bel
lsou
th
Bel
lSou
th
PacT
el
Puer
to R
ico
san Juan
NA
N
A
Artansas
Littl
e R
ock
sw Bel
l B
ellS
outh
Pa
cTel
sw
Bell
Michigan
Sagi
naw
A
mer
itech
A
mer
itech
B
ells
outh
Pa
cTel
So
uth
Car
olin
a C
olum
bia
Bel
lSou
th
Bel
lSou
th
PacT
el
Cal
iforn
ia
Bak
ersf
ield
P8
cTel
Be
llSoU
th
PacT
el
sw Bell
Ncv
ada
Las
Veg
as
PacT
el
Bel
lSou
th
PacT
el
sw Bel
l
Indi
ana
Fort W
ayne
A
mer
itech
A
mer
itech
B
ellS
outh
Iow
a D
aven
port
us west
Am
erite
ch
Bel
lsou
th
Penn
sylv
ania
Y
ork
Bell A
tlant
ic
Bel
lSou
th
sw Bel
l Lo
uisi
ana
Shre
vepo
rt B
ells
outh
B
ellS
outh
Pa
cTel
sw
Bell
TCXaS
Be
aum
ont
sw Bel
l BellSouth
sw Bell
Iow
a D
es Moines
us W
est
Bel
lSou
th
sw Bell
us W
est
lllin
0i.S
Pe
oria
A
mer
itech
A
mer
itech
B
ellS
outh
V
irgin
ia
New
port
New
s B
ell A
tlant
ic
Bell A
tlant
ic
Bel
lSou
th
PacT
el
Penn
sylv
ania
La
ncas
ter
Bell
Atla
ntic
B
ellS
outh
sw
Bell
us W
est
us W
est
Sow
er: B
IA P
ublic
atio
ns (1
993)
. N
ore:
The
sale
of S
outh
wes
tern
Bell's
prim
ary
pagi
ng s
ervi
ce s
ubsi
diar
y, M
etro
med
ia P
agin
g Se
rvic
es, a
nd th
e sa
le o
f Pa
cTel
's pr
imar
y pa
ging
se
rvic
e su
bsid
iary
, Pac
Tel P
agin
g, a
re pending.
THE EMERGENCE OF COMPETITION AMONG THE BABY BELLS 487
Table 5. RBOC versus Competition in Yellow Pages Publishing within Selected Metropolitan Statistical Areas, 1993
Ameritech
BellSouth
"Ex
Pacific Telesis
Southwestern Bell
Fort Wayne, IN
Indianapolis, IN Dayton, OH
Chicago, IL
Louisville, KY-OH Johnson City, TN-VA Tampa, FL Memphis, TN-AR-MS Birmingham, AL Mobile, AL
Bridgeport, CT Albany, NY New York, NY-NJ
Los Angela, CA San Diego, CA San Jod, CA Sacramento, CA h a r d , CA Las Vegas, NV S.F.-Oaklad, CA
St. Louis, MO-IL Oklahoma City, OK Beaumont, TX
Houston, TX Dallas-Ft. Worth, TX
Southwestern Bell us West us west us west
Ameritech Bell Atlantic Southwestern Bell Southwestern Bell Southwestern Bell Southwestern Bell
us west us West us west, Bell Atlantic
" E X "Ex "Ex "Ex "Ex Southwestern Bell us West
Ameritech us west us west US West US West
US West Davenport, IA-IL Ameritech Omaha, NE-JA Southwestem Bell
Note: Metropolitan areas are selected from the 105 largest Metropolitan Statistical Areas (MSAs) in 1980. Some: Yellow Pages Publishers Association (1993).
instability in the RBOCs' market shares over time is at odds with the possibility of a collusive agree- ment in cellular services. Rather, the evidence is consistent with the existence of spirited competi- tion between the RBOCs in this market.
Regression Analysis
The MFTs assumption that the RBOCs will col- lude also can be tested by examining the prices charged in markets where they compete head-to- head. If the RBOCs were colluding successfully, then prices should be higher in those geographic areas that are served by two or more of them, as compared to areas served by a single RBOC and
a non-RBOC competitor or competitors. This possibility was examined empirically in two cases: cellular telephone services and yellow pages pub- lishing.
The local RBOC's rate for cellular telephone services in cities where RBOCs compete head- to-head was compared to the rate in cities where the local RBOC faces a non-RBOC competitor. Regression analysis was used to test whether rates were signi6cantly higher in markets where two RBOCs compete. The dependent variable was the home RBOC's local wireline rate (CYPRICE) in 125 metropolitan statistical areas (MSAs). The local wireline rates were regressed on a binary variable (BOC) which assumed the value of 1 if
488 WILLIAM F. SHUGHART I1
25.096
lmD%
H 15.096 a %! I-
i 5.096
0.096
-
-
-
: -
-
........................................................................................................................
.................................................................................................
...................................................................................
Figure 1. Individual RBOC shares of total cellular subscribers, 1988 and 1992. Source: EMCI 1993, p. 96 .
there was RBOC versus RBOC competition in variable definitions and data sources is provided the MSA, the population density of the MSA in Table 6.) (DENSZW), and the per capita income in the The population density variable was included to MSA (PCZNCOM.). (A complete listing of the control for variations in the cost of providing
Table 6. Variable Definitions: Cross-Section Analysis of Cellular Telephone Rates
Variable kenition and source
W R K E The real RBOC local wireline rate based on a 150-minute customer bill. Deflated by the relevant city or regional all-item CPI-U (Paul Kagan Associates, 1992 pp. 9-14)
DENSITY The MSA population density per square mile. Calculated from MSA population and land area figures. (US Bureau of the Census, PoprJarion of M e ~ p h Anac ond Componenf Geogmph: 1990 CPH-Ll45, data diskette)
PCINCOME The real per capita income in the MSA. Defhted by the relevant city or regional all-item CPI-U. (US Bureau of the Census, Income Swnmary Meanucs by Mebopoliran StatiaicalAnac, data diskette)
A binary variable which assumes the value of 1 if there was RBOC versus RBOC competition in the MSA, and 0 otherwise
BOC
THE EMERGENCE OF COMPETITION AMONG THE BABY BELLS 489
cellular telephone service across cities. It is ex- pected that the cost of providing cellular service will be higher in more densely populated areas because of the need for more cells in areas of intense cellular use. Per capita income was in- cluded in the regression to hold constant varia- tions in the demand for cellular service across cities, with the expectation that higher income areas will have a greater demand for cellular telephone services. The binary variable, BOC, was included to test for evidence of collusion in the cities where cellular service is offered by more than one RBOC. The estimated coefficient on this variable will be positive and significant if the RBOCs have cooperated in establishing the prices of cellular telephone services.
The results of the regression analysis shown below give no evidence of collusion in the setting of wireline rates (absolute values of t-statistics in parentheses):
WRICE = 40.85 + 0.003 DENSZW (1.97)
+ 0.002 PCZNCOME
- 9.017 BOC (2.44)*
(4.31)** R2 = 0.18, F = 8.996**, N = 125
(Asterisks denote significance at the 1% ?*) and 5% P) levels.)
In particular, the coefficient on BOC is nega- tive and significant at the 1% level, indicating that the local wireline rate is actually lower, not higher, in cities served by more than one RBOC.
The sign and significance of the remaining co- efficients are consistent with the predictions of economic theory. In particular, the coefficients on both population density (DENSITY) and per capita income (PCZNCOME) have the expected positive sign, suggesting that wireline rates for cellular telephone services are higher in more densely populated metropolitan areas and in ar- eas where incomes are higher.I3
In short, the empirical evidence weighs against any possibility of RBOC collusion in the case of cellular telephone services. Wireline rates tend to be lower in markets served by more than one RBOC and, when compared with the rates charged in markets served by only one RBOC, the difference is statistically significant. (Figure 2
compares the average wireline rates charged for cellular telephone services in the two types of markets.)
Cross-section regression analysis was also used to test for evidence of collusion in yellow pages publishing. Five different yellow pages listing rates charged by the home region RBOC in cities where it did compete against another RBOC were com- pared with those in which it did not. The five yellow pages listing rates examined where regular listings (RL), bold name and number listings (BLN), double half-column displays (DHC), dou- ble quarter-column displays (DQC), and quarter- column displays (QC). The rates were regressed on a binary variable (BOC) which assumed the value of 1 if there was RBOC versus RBOC competition in the city, city population (POP), and per capita income (PCZNCOME). (A com- plete list of the variable definitions and data sources is provided in Table 7.)
Both population and income were included in the regression to control for variations in the demand for advertising across cities. Both of these variables are proxies for the size of the market, and one would expect the demand for yellow pages advertising to be greater in larger markets. Cost variables were not included in the regression specification on the assumption that the cost of publishing yellow pages does not vary from city to city. Once again, BOC is included to test for evidence of collusion. If the RBOCs have cooper- ated in the setting of the rates charged for yellow pages advertising, the estimated coefficient on this binary variable will be positive and signifi- cant.
The results of the regression analysis reported in Table 8 are inconsistent with the collusion hypothesis. The estimated coefficient on BOC is not statistically different from zero in any of the specifications. BOC’s lack of statistical signifi- cance suggests that the rates charged for yellow pages listings are no higher in cities served by two or more RBOCs than in cities served by only one of them.
The sign and significance of the coefficients on the remaining variables are again generally con- sistent with the predictions of economic theory. The estimated coefficients on both population (POP) and per capita income (PCZNCOME) are positive in four out of the five regressions. And, except for per capita income in the bold name
490 WILLIAM F. SHUGHART Jl
$100
................................................................................................................................................................................................................................................................#� F @o .............................................................................................. L ’i 520 so
.....................................................................................
.................................................................. ...................................... ......................
Figure 2. RBOC mean real wireline price, RBOC non-wireline competitor versus other non-wireline competitors, 1992. (Source: Paul Kagan Associates 1992, pp. 9-14.)
Table
Variable
RL
BLN
DHC
QC
DQC
BOC
POP
PCIN- COME
7. Variable Definitions: Cross-Section Analysis of Yellow Page Advertising Rates
Definition and source
Real rate charged by the home region RBOC for a ‘regular’ yellow page listing. Deflated by the relevant city or regional all-item CPI-U. (YPPA, 1993a, pp. 389-559)
Real rate charged by the home region RBOC for a bold name and number listing. Deflated by the relevant city or regional all-item CPI-U. (YPPA, 1993a, pp. 389-559)
Real rate charged by the home region RBOC for a double half column listing. Deflated by the relevant city or regional all-item CPI-U. (YPPA, 1993a, pp. 576-745)
Real rate charged by the home region RBOC for a quartercolumn listing. Deflated by the relevant city or regional all-item CPI-U. (YPPA, 1993a, pp. 576-745)
Real rate charged by the home region RBOC for a double quarter-column listing. Deflated by the relevant city or regional all-item CPI-U. (YPPA, 1993a, pp. 576-745)
Binary variable that assumes the value of 1 if there was RBOC versus RBOC competition in the city, and 0 otherwise
The population of the MSA. (US Bureau of the Census, Population of MempolsOn Areas and Component Geogmphy: 1990 CPH-G145, data diskette)
The real per capita income in the MSA Deflated by the relevant city or regional all-item CPI-U. (US Bureau of the Census, Income Summuy Meanues @ Mehvpolitan Statisiical h a s , data diskette)
THE EMERGENCE OF COMPETITION AMONG THE BABY BELLS 491
Table 8. Regression Results: Yellow Pages Advertising Rates
RL BLN DHC OPC Constant 100.24 138.51 1970.85 1109.24
POP - 8.94E - 07 1.32E - 05 4.92E - 04 2.39E - 04 (0.25) (3.28)' (3.73)' (3.49)'
PCINCOME - 1.40E - 04 0.003 0.262 0.123 (0.04) (0.85) (1.99)C (1.80)E
BOC - 5.68 - 9.38 - 89.46 - 60.565 (0.46) (0.66) (0.19) (0.25)
R2 0.01 0.17 0.27 0.24
F 0.14 5.40a 9.87' 8.43'
N 65 85 86 86
ec 758.73
1.22E - 04 (2.99)'
0.051 (1.34)
2.227 (0.02)
0.18
5.77a
83 ~~ ~~
Notes: Absolute values of &statistics in parentheses. 'Significant at the 1% level. bSigniiicant at the 5% level. 'Significant at the 10% level.
and number listing (BLN) and quarter column (QC) listing regressions, the coefficients on both of the variables are sigdicant at the 10 percent level or better in these four cases.14
Acknowledgements The author wishes to thank Michael McDonald of Capital Economics for able research assistance. The usual caveat applies.
NOTES CONCLUSION
Regulatory and market changes since 1984 ensure that, if allowed to enter into manufacturing and interexchange markets, the RBOCs will not en- gage in the cross-subsidization or foreclosure that was feared at divestiture. Nor is there any reason to fear collusion among them. Descriptive and econometric evidence indicates that the RBOCs have competed with one another at every oppor- tunity since divestiture. The RBOCs' unwilling- ness or inability to collude in the past makes it unlikely that they will do so in the rapidly chang- ing technological environment of the future.
There is every reason to believe that the entry of the RBOCs into equipment manufacturing, interexchange service, and other adjacent markets will increase output and enhance the overall com- petitiveness of the telecommunications industry. I therefore strongly support the lifting of the line- of-business restrictions in the Modification of Fi- nal Judgment.
1. United States u. American Tel. & Tel. Co., 552 F. Supp. 131, 227-28 (DDC 19821, a f d sub nom. Maryland u. United States, 460 US 1001 (1983).
2. See id. at 190. 3. United States u. Western Elec. Co., 673 F. Supp. 525,
558 (DDC 1987), a f d in part and reu'd in part, 900 F.2d 283 (DC Cir.), cert. denied, 498 US 911 (1990).
4. Id. 5 . 15 USC 5 15. 6. 15 USC $8 1, 2. 7. See e.g., United States u. Pippen, 903 F.2d 1478,
8. See generally Shughart (1990, pp. 225-238). 9. 673 F. Supp. at 558. But see 900 F.2d at 296 ("IlIe
district judge's speculation that the BOG could impede competition by way of illegal (and perhaps criminal) collusion to divide markets among them according to territory would, in the absence of supporting evidence, seem to quallfy only as a theoretical possibility') (citation omitted).
10. For example, McChesney (1995) finds that GTE, an independent telco equivalent in sue to an individ- ual RBOC, was unable either to exclude or dis- criminate against rivals in the interexchange mar- ket or to successfully deter new entry.
1482-84 (11th Cir. 1990).
492 WILLIAM F. SHUGHART Il
11. The following RBOC-cable TV alliances were an- nounced in 1993: Bell Atlantic's attempted merger with Tele-Communications, Inc.; BellSouth's pur- chase of 22.5% of Prime Cable; " E x ' s an- nounced investment of $1.2 billion inViacom Inter- national; Pacific Telesis' option to acquire 75% of Prime Cable; Southwestern Bell's purchase of two cable systems in Virginia and Maryland from Hauser cable and its attempted joint venture with Cox Enterprises; and US West's purchase of a 25% stake in Time Warner Entertainment Co. See Broadcasting, 15 February 1993 at 3; Broodcarting & Cable, 24 May 1993 at 6; Wall Street Jownal, 5 October 1993 at A3; The (Montreal) Gazette, 30 October 1993 at D1; New York Times, 28 Novem- ber 1993 at section 3, p.1; Wall Street Journal, 7 December 1993 at A3.
12. Besides providing one more example of competi- tion among RBOCs, the RBOC-cable TV alliances are noteworthy because they further diminish the RBOCs' ability to collude. Such alliances speed the elimination of any 'bottleneck' at the local tele- phone exchange. Cable alliances give competing RBOCs the ability to provide local customers ac- cess to long-distance and on-line data services, bypassing the home RBOC's local exchange net- work. Bypass of the local exchange renders ob- solete the MFJ's basic premise that RBOCs could abuse their control over monopolistic local ex- change facilities.
13. Two other specifications of the regression equation were estimated. When the population of the MSA was included, the coefficient on BOC remained negative and significant. In addition, dummy vari- ables for each state were added to the specification as a way of controlling for differences in the local regulatory environment. While the estimated co- efficient on the BOC binary variable was positive in this case, it was not different from zero at standard levels of statistical significance. These re- sults are available on request.
14. These results are not sensitive to taking account of the actual number of competing yellow page direc- tories available in a given metropolitan area (OTHCOMP). When OTHCOMP was included in
any of the five regression specifications, the esti- mated coefficient on BOC was consistently insig- nificant. (The estimated coefficient on OTHCOMP was itself negative and insignificant in all live speci- fications.) Moreover, there is no statistically sig- nificant difference between the home RBOC's ad- vertisiig rates in cities where out-of-region RBOCs offer yellow pages and the rates charged by the home RBOC in other cities. This result is also not affected by the total number of competing yellow pages directories available in a particular geo- graphic market. This analysis provides further evi- dence that the RBOCs are competing rather than colluding in the yellow pages market.
REFERENCES
BIA Publications (1993). The 1993 Pa@g/PCS Direc- tory, Chantilly, VA BIA Publications, Inc.
Cellular Telecommunications Industry Association (1992). State of the Cellular Zndushy, Washington, DC.
Economic and Management Consultants International, Inc. (EMCI) (1993). The Cellulur Marketplace: 1993, Washington, DC.
Paul h g a n Associates, Inc. (1992). Cellular Rates 1992, March, Carmel, CA.
F. S. McChesney (1995). Empirical tests of the cross- subsidy and discriminatory access hypotheses in ver- tically integrated telephony. Managerial and Deciswn Economics, this issue.
W. F. Shughart I1 (1990). The Organization of Zndzuhy, Homewood, 11: Richard D. Irwin.
G. J. Stigler (1964). A theory of oligopoly. Jownal of Political Economy, 72, February, 44-61.
US Federal Communications Commission (1993). Mon- itoring Report, Staff of the Federal-State Joint Board, Washington, DC, May.
US Telephone Association (USTA) (1993). 1993 Phone Facts, Washington, DC.
Yellow Pages Publishers Association (YPPA) (1993a). Rates & Data Part I , , Vol. 1, October, Troy, MI.
Yellow Pages Publishers Association (YPPA) (1993b). Rates t Data Part 2, Vol. 2, October, Troy, MI.