FTC Staff Comment Before the Federal Energy Regulatory Commission Concerning Inquiry Into Alleged Anticompetitive Practices Related to Marketing Affiliates of Interstate Pipelines

• • }....1
,-' 	/ 	/~s C 	.
COMMISSION· ~
APPROVED
- .~ .. ~~ 	,~.-.- .,-
UNITED STATIS OF AMERICA
FEDERAL ENERGY REGULATORY COMMISSION
IDqulry lato AJl~led ADtlcolDpetlth~
Practices Related to MarkttlDI AffiUates
or IDterstate PlpeJlnu
)
)
) 	Docket No. RM87-S-000
ST ATIME:~T OF THE: UNITED STATES FEDERAL TRADE COMMISSION ST AfF
January 	29, 	1987

Table of eo.teats
L berMuctioll 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	• . 	. 	.
IL The Reculatloll of latenute Gas PipeUDeI
A. The Curreat R...ulatory E •• lro.....t .
1
B. Potntlal Proble.s WIth the Currnt aeculatory E •• iro ••nt 	7
1. Third Party Allelatlolll of MoaoplollY Power. 	. 	. 	. 	. 	. 	B
2. Third Party Allelatlolls of Dow.seream MODOpoly Power 	10
3. PlpeUDe Behulor Tbat MiChl Not be A.tlumpetlthe . 	11
Ill. 	FER C's Policy Alter..thes. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	. 	13
A. FIRe Should Eac:ourale Maximum PlpellDe CompetltloD . 	13
B. Reculatory Alter.athes 	.
1. ReculatioD of MarkulDI AfflUaUs
2. 	Dhestlture of MarketiDC Affiliates
3. 	ProblbitiDI the MarkeHDI Firms From UdDI Their
Affiliated 'Iptlloes
~. RtcommudatloD
IV. CODdusioD
Appeadix ...
16
18
20
23
2S
26
29

I
UNITED STATES OF AMERICA
FEDtRAL INtRGY REGULATORY COMMlSSIOl\
laqulry lato Alleled ADtlcolDpetlthe
Practices Related to MarketlDI Afflilatel
pr btustlte Pipelines
)
)
) 	Docket No. RMI7·5-000
StatelDeDt of tbe F.deral Trade COlDlDlssloD Staff l
I. btroductioD
The 	staffs 	of 	the 	Bureaus 	of Competition, Consumer 	Protection, and
Economics of 	the Federal Trade Commission appreciate this o~portunity to
res~ond to the 	Federal 	Energy 	Regulatory Commission's (FERC) ·Notice of
Inquiry (NOn' into alleled Inticompetitive practices related to the marketing
affiliates 3 of interstate pipelines. The specific purpose of the inquiry is to
determine 	whether 	vertical 	intearation 	by 	pipelines 	into 	the 	marketing 	of
na tural 	las 	raises 	the 	costs 	of 	rival 	marketers 	that 	are 	not 	in tean ted,
thereby 	reducing 	competition. 	FIRC 	also 	seeks 	SUllestions 	on 	possible
regulatory 	res~onses if the evidence indicates that anticompetitive behavior
exists.
1 	This statement re~resents the views of the staff of the Federal Trade
Commission. 	It 	does 	not 	necessarily 	represent 	the 	views 	of 	the 	Federal
Trade Commission or of any individual Commissioner. 	The Commission has.
bowever, authorized the staff to submit this statement.
2 	Federal EnerlY Regulatory Commission, NOliu 01 Inquiry into Alltgtd
A"'ico,"~'i'ivt 	Praclices 	ReJaud 	to 	Marktli"g 	AfliJialts 	of 	1"ltrSlalt
Pi~Jillts. U CFR Parts 154 and 271, November 14, 1986.
S 	A 	-marketing 	affiliate- 	of 	an 	interstate 	pipeline 	is 	essentially 	a
broker of natural las that is owned by the pipeline. 	That U, it enters into
eOrltncts both to buy 	gas from 	producers (who may be independent of the
pipeline), and to sell las to customers, such as industrial end-users Ind local
distribution 	companies. 	It 	may also arrange for 	the 	,as to be transported
from producers to customers, using both affiliated and unaffiliated pipelines.

Thc fedcral Tradc Commission (Commission) is an indcpe~~cnt Ilcncy
with 	thc 	responsibility 	tor 	cnforc:inl 	the 	fcdcral 	Trade 	Com:ission 	Act,·
which prohibits ·unfair mcthods of competition.· 	The Commission seeks to
promotc 	compctition 	in 	cncrlY 	markcts, 	IS 	cvidcnccd 	by 	II Ulcrous 	law
cnforcement actions and policy statcmcnu beforc various relulatory bodics,
inc:ludinl FERC.' 	The Commission's naff hiS IcncralJy espouse~ thc position
that thc historic practice of rClulatin& price and cntry bchavi::- in natural
las production and distribution bas hindcrcd the ability of tha: industry to
./
respond 	to 	market 	cbanaes. 	to 	the 	dctriment 	of 	consu~::, 	welfare.'
Consequcntly, we have uraed that fIRC adopt policies that pll:e incrcased
rcliancc 	on 	competitivc (orces, 	whcnever 	thesc 	policies are co~istcnt both
with FIRe's statutory obliaations and with undcrlying markct cbractcristics.
The 	NOI sceks information on 	thirty 	spccific issucs dceced 	to 	be of
rclevance 	to 	the 	potcntial 	(or 	anticompctitivc 	behavior 	b~ 	iDtcluted
pipeline/markcters. 	For 	tbe most 	part, wc 	do not attcmpt 	te answcr each
Question individually, but rathcr direct our response to thc brOtd structural
and cconomic issues that cover many of thc Questions.
These 	comments 	are 	oralnized 	as 	toUows. 	Section 	II 	identifies 	the
potential 	anticompetitive 	problems 	that 	miaht 	arise 	under 	the 	current
regulatory environment. and discusscs lome instances wbcrc atticompetitive
activities 	are 	aUcaed 	to 	bave 	takcn 	place. 	Section 	III 	disnsses 	FIRC's
policy 	alternatives. 	Tbese 	consist 	of 	(I) 	continuing 	to 	all:... 	interstate
4 	1S U.s.C. 41 tl Stq.
. 	' 	Sec 	StaUmt1l1 01 tht 	Ftdtral Tradt 	Commissioll 	Stalf. Inquiry 	Into
~urchasing Practices 	of 	Interstate 	Pipelines, 	FIRC 	Okt. 	No. RPS3·96·000,
luly I, 1983.
• 	Id.
:2

pipelines to own and operate unrelulated maiketinl affiliates. (2) relulatin,
the prices charled by marketinl affiliates, (3) requirina pipeliDes to divest
their 	marketinB 	affiliates. 	and 	(~) restrictin, 	the 	access 	of 	marketers 	to
their 	affiliated 	pipelines. 	Section 	IV 	is 	the 	conclusion. 	M 	appendix
provides responses to some of the specific questions posed ill tbe NOL'
Sued upon theory and available evidence. we believe tbat there may be
instances 	wbere 	integrated 	pipeline/marketers can 	bebave anticompetitively
towards rival marketers. and 	tbereby increase their own profits. 	However,
we find it plausible (thou,h by no means proven) that tbere may be lenuine
efficiencies 	Benerated 	by 	the 	integration 	of 	,as 	transmission 	with 	las
marketing. 	Selection of the optimal policy therefore involves balancing tbese
efficiencies 	against 	the 	potential 	welfare 	losses 	created 	by IDticompetitive
bebavior. 	We counsel alainst attempting to regulate the prices Ind marlins
received by marketing affiliates. 	The bistorical experience with wellhead ,as
pricing 	regulation 	SUllests 	that 	such 	regulation 	is 	inefficient 	and
distortionary. 	Instead, 	we 	recommend 	tbat 	FERC live 	serious 	tbought 	to
adopting 	a 	·no access· 	policy 	wbich 	prohibits affiliates 	from 	shipping 	on
tbeir 	parent 	pipeline. 	The 	principal 	virtue 	of 	tbis 	straten 	lies 	in 	its
flexibility. 	If 	the 	cost 	of 	forelone 	vertical 	efficiencies 	appears 	to 	be
Ireater 	than 	orilinatly 	anticipated, 	FERC 	can 	remove 	tbe 	restriction 	and
permit marketers to purcbase .pace from tbeir affiliated 	pipelines. 	If the
policy 	proves 	to 	be 	ineffective 	as 	a 	means 	for 	deterring 	Inticompetitive
bebavior, FERC can then require outrilht vertical divestiture. 	Moreover, as
a 	first 	step, 	FERC sbould 	consider 	strenathening 	its 	·open 	access· policy
(i.~ whereby pipelines sell space to independent shippers on • (irst come,
.. T 	We respond to questions I, 2, 4, 6, 1S, and 21.
3

Cint served 	basis) by 	requirinl 	pipelines 	to ,ive advanced 	Dotice of nrm
transportation 	capacity, 	and 	by 	adoptinl 	policies 	allowin, 	the 	ri,bt 	to
interruptible 	Ipace 	to 	be 	vested 	in 	'IS 	purchasers. 	We 	believe 	that 	the
adoption 	of 	these 	policies 	will 	ultimately 	result 	in 	the 	enhancement 	of
consumer welfare.
II. Tbe RelulltloD or I.ttrUlt. Cas Plp.llus
A. The Currnt Relulltory EDl'lrODDIeat
To 	understand 	the 	competitive 	issues 	beinl 	raised 	in 	the 	NOI, 	it 	is
useful 	to 	visualize 	the 	lIS 	industry 	as 	consistinl 	of 	thre: 	Ita.es: 	(J)
production, (2) transportation, and (3) local distribution. 	In addition, there
are 	firms 	which perform' DODe of these three CuDctions, but which instead
simply 	buy 	aDd 	sell 	IU 	OD 	the 	open 	market. 	These 	nrms 	are 	typically
referred 	to as -brokers- or -marketers.- 	Natural las producers explore for
,as, develop reserves, produce, and ultimately lell las. 	Historically, most of
this 	las 	has 	been 	lold 	to 	pipelines, 	who 	transport 	the 	las 	to consuminl
areas 	aDd 	deliver 	lIS 	either 	to 	IUle 	volume 	direct 	users 	(C.I... 	electric
utilities), or to local distribution companies (LDCs). 	The LDu then deliver
the .as to small volume users, such as households. 	All 'IS lnDuctions do
not follow this pattern, however. 	As an alternative, an LDC mi,ht purchase
,as 	from 	a 	broker, 	who 	has already 	arranled 	to purchase 	,as Crom some
source, such as a producer.1 	The broker miaht then urlnle {or a pipeline
to deliver the aas to the customer.
" 	I 	The broker has other sources of las besides producers; itmi,ht, for
example, purchase las from I 	pipeline that wishes to dispose of excess liS
in ven [or ies.

The 	Natural 	Gas 	Policy 	Act 	of 	1971 	(NGPA)t essentiall)' eliminated
Cederal 	control 	of wellbead 	las prices,10 as 	well as controls over cntry or
exit 	into aas 	production or exploration. 	PassaiC of the 	NGPA constituted
Conlressional 	reco,nition 	of 	the 	competitive 	nature 	of 	the 	IU production
and exploration industry.
On 	the 	otber 	band. 	because 	IU 	transportation 	is 	cbara::terized 	by
increasing returns to scale. individual pipelines in some aeolraphic markets
may be -natural monopolies.- 	Thus. it is not uncommon to find producers. as
well as local distribution companies (LDCs) and end-users, servee: by a single
pipeline. 	This creates two opportunities for the exercise of market power by
pipelines. 	First, 	they 	may be able 	to act as 	monopsonisu with respect 	to
las producers, which 	means that they will 	pay a sub-competitive price for
wellhead 	.as, 	and 	second, 	they 	maybe 	able 	to 	let 	as 	monopolists 	with
respect to LDCs and end-users. which means that they will rese:l the .as at
above~ompetitive prices. 	Because 	of 	this 	potential 	for 	anticompetitive
behavior. Congress elected to retain controls over pricing. entry. and exit in
the 	market 	for aas tnnsmission services. 	OwinB to the nature of the cost
and demand conditions cbaracterizinB this service. it was felt that (at least
in certain 	geographic markets) there existed I 	potential for anticompetitive
market 	behavior. 	Accordingly. interstate 	.IS 	pipelincs arc stilI subjcct 	to
FERC regulations on pricinl and investmcnt dccisions.
t 	15 U.s.C. 717.
10 	Some eatcaorics of las were not decontrolled under tbe terms of the
NGPA. 	However, FERC Order No. "51 	(issued June 6. 1986) establishcs •
pricing 	structure 	approximating 	that 	which 	would 	exist 	1lDder 	actual
competitive conditions.
.5

In receDt yean, FERC has attempted to develop policies that will allow
las producers and consumers to take advanlaae of the competitive Dlture of
the 	Datural 	las 	market. 	while 	simultaDeously 	constrainiD, 	pipcliJles 	Crom
cxercisin, 	market 	po....er. 	The recently-promulaated 	FERC Order 	No. ~36,
which established aD ·opeD Iccess· policy, is the most promineDt example or
this cDdeavor.ll 	MaDY iDterstate pipelines DO.... eDter iDto CODtT'l:U to Clrry
IU 	that 	is owned 	by 	others. 12 	Order 	No. 4C36 requires 	that all 	iDterstate
pipelines that hold themselves out as transporters of las for oth::-s do so on
a 	Dondiscriminatory basis. 	This melns that the 	pipeline is required to sell
space at 	(or below) the 	FERC tariff on a 	first-come, 	first-served 	basis to
any 	shipper 	(e.I .. 	I 	producer, 	an 	LOC, 	In 	end-user, 	aD 	iDdependent 	las
broker, or a marketinl a(fililte) who ....ishes to buy it. 	Participation in the
Order No. ~36 prOlram is voluntary. 	If I 	pipeline elecu to participate in
the 	program, FERe will 	issue a 	blanket certificate 	for 	this 	transportation
service. 	thereby 	reducing 	reaulatory 	delays 	or 	impedime~!.S 	to 	these
traDsactions. 	By separating the 	purchase of lIS from 	the 	purehase or ,as
transportation services, Order No. 4C36 may provide buyers and seIlen of las
....ith 	access 	to 	a 	much 	Ireater 	Dumber 	of 	potential 	trading 	partners 	than
they had heretofore enjoyed.
11 	See FERC Order No. 4436. Rtgulatioll of Natural Gas P,~li1US Aflt'
Partial Wtllhtad DtCOlltrol. Okt. No. RM85-1-o00, October 9. 1985.
.	t.
12 	Traditionally, pipelines transported only their own gas.
6

B. Poteatlal Proltle•• Wltb tbe enrt.t Rfellatory [ulro••ut
As 	discussed 	above, 	Coneress 	bas 	rccolnized 	that 	pipelines 	can
sometimes exercise market power. and bave therefore impos~ relulations on
the 	prices 	charled 	when 	pipelines 	<as 	opposed 	to 	pipelincs' 	marketing
affiliates) 	·sell 	1&5 for 	resale· 	<e.,., sell 	las 	to 	LDCs), as 	well 	as 	on 	the
tariffs 	charled 	when 	pipelines 	transport 	las 	for 	others. 	FERC 	has 	also
established ·open access· policies. 	If effective, tbese relulations reduce both
monopoly 	and 	monopsony power. 13 	Some opportunities for Inti:ompetitive
behavior 	may 	nonetheless 	remain. 	Many 	pipelines 	own. or 	arc 	affiliated
with. firms that buy and sell las on the open market. 	At the present time.
the 	prices 	charged 	by 	these 	marketing 	affiliates 	are 	not subje:: to 	FIRC
regula tion. 14
The ability of a pipeline to establish an unrelulated marketing affiliate
may create In opportunity for the pipeline to earn supracompetitive returns.
If an ·open Iccess· pipeline can make it costly or difficult for iDdependent
shippers to purchase space, it may provide its marketing Iffilia:: with the
ability 	to 	behave 	in 	In Inticompetitive 	Cashion. 	A captive lIS producer
(i.e, 	I 	producer 	served 	by 	only 	one 	pipeline) 	which 	might otherwise 	have
executed a competitive sales contract directly with an LDC. lIli&~t have 	no
13 	In 	theory, 	the 	rate-of -return 	regulation 	reduces 	the 	pipeline's
incentive 	to 	Ict 	IS 	I 	monopsonist. 	since 	the 	regulated 	price at 	which 	it
resells the 	gas will 	be adjusted downward 	to reflect Iny reductions in 	the
price 	at 	which 	it 	Icquires 	the 	las. 	However, 	IS 	we 	explain 	below, 	if 	a
pipeline can establish an unregulated marketing affiliate, monopsony behavior
may once again become profitable.
14 	For 	example, Tenneco 	Corporation 	owns 	both 	tbc- Tennessee 	Gas
Pipeline Company and the Tenngasco Corporation. 	Tenngaseo is • marketer;
it buys and sells las on the open market. 	Wben Tennessee Pipeline lells ,as
Je 'an LDC. the price is regulated; when Tennaasco makes I 	simlar sale, tbe
price is market-determined.
7

alternative but to sell to a marketin,arfiliate '(at a monopsony price). if the
marketin, affiliate has acquired all o( the available transportation space. 	An
LDC. if not served by other pipelines. mi,bt likewise find itself with little
alternative 	but 	to 	buy 	lIS 	from 	a marketina affiliate 	of a 	pipeline (at a
monopoly 	price) 	if 	otber 	potential 	las 	suppliers 	cannot 	obtain 	space 	at
competitive terms. 	Lastly. the existence of an 	unreaulated subsidiary may
permit 	the 	parent 	company 	to 	cnsale 	in 	profitable 	cross-1ubsidintion
between 	subsidiaries. 	In 	situations 	wbere 	one 	subsidiary 	is subject 	to 	a
binding profit constraint. it may sometimes be profitable (or the parent to
csta blish an a(filia te tbat will sell in an unrelulated market. 	If some of the
unreaulated 	cntity's costs can 	be incorporated 	into the reauJated affiliate's
rate base, tbe resultinl hilber tariff (and revenue) may increase the firm's
overaU 	level 	o( 	profitability. 	even 	in 	situations 	wbere 	tbe 	unregulated
subsidiary cannot cover all of its costs at tbe competitive price.
1. Third Part)' AlleaatloDs of MODOPIOD)' Power
A number of tbird parties have submitted documents to FERC in which
tbey discuss possible instances of monopsonistic behavior. 	In a'recent FERC
proceeding,1I 	Champlin Petroleum Company (an oil and 	,as producer) has
contended 	that 	marketing affiliates can obtain superior access to 	pipelines
through 	anticompetitive 	means. 	Tbis allows 	them to extract -Unreasonable
marketing 	fees· 	from 	producers. 	Since 	these 	fees 	reduce 	the 	net 	price
received 	by 	las 	producers. 	this 	is 	simply 	another 	way 	of 	stating 	that
marketing affiliates wiU exercise monopsony power llainst their suppliers.
to 	11 	See 	FERC 	Order 	Stuin, 	Ctrti/ica!t 	Application 	lor 	Htaring.
Tenngasco 	and 	TennllSco 	Exchange 	Corporation, 	Dkt. 	No. 	0&6-168-000,
September 11, 1986.

In 	a 	formal, complaint 	before 	FERC, 	thelndependt:t 	Producers
Association 	of 	Mountain 	States 	(IPAMS) 	has 	aIle led 	that 	two 	pipelines,
Panhandle 	Eastern 	Pipeline 	Co. 	(PEPl) 	and 	Colorado 	Inttntltc 	Gas 	Co.
(00), refused 	to disclose 	to it accurate transportation 	rates. ud available
capacities, for shipments of las 	to end-usc markets that tPAMS desired 	to
serve. 11 	Accordin, to the complaint, this behavior prevented IPAMS (or any
other nonaffiliated shipper) from arranlinl sales to these markets. 	This, in
turn, allowed the pipelines' marketing affiliates to execute sale3 contracts to
most of the LDCs and end-users in these markets." The ina biliry of IPAMS
to sell las directly to end-users and LOCs not only created opportunities for
the exercise of market power by pipelines in downstream markeu,11 but also
(or 	the 	exercise 	of 	monopsony 	power 	in 	upstream 	markets. 	If 	producers
such 	as 	IPAMS are 	not served 	by 	other 	pipelines, and canno: arran Ie 	for
competitively priced 	transportation of las for 	direct sales to end-users and
LOCs, 	then 	their 	sales 	opportunities 	may 	be 	restricted 	to 	the 	pipeline's
marketing affiliates. 	The result will be, as IPAMS contends in its complaint,
below-competitive wellhead prices for las.
11 	IndependeDt 	Producers 	Assn. 	of 	Mountain 	States 	v. 	Panhandle
Eastern Pipeline Co., Okt. No. CP86-S84-000, June 20, 1986 .
•' 	11 	The existence of tbese opportunities will depend upon the Dumber of
alternative fuel sources that are available to PEPL's and CIO's customers.
9

1. Tblrd Part, Allt.atloa. of Dowa.tream Moaopol,Pow"
Several petitionen blve described methods by which pipelines can raise
the 	costs 	of 	their 	rival 	marketers. 	Hadson 	Gas 	Systems1... 	I 	producer,
marketer, and 	transporter of 	lIS, bas claimed 	that 	pipelines have imposed
lupncompetitive -imbalancin, penalties· on indepcndent shippen. 1t 	Althouah
the marketinl affiliates arc subject to the Slme penalties, Hadson claims that
thc 	affiliatcs 	are 	Dot 	disadvantaged 	by 	tbis, 	since 	the 	payment 	of 	the
penalty is a pure intrafirm transfer.
Hadson's claim may not be 	valid. 	First, if the 	penalties only reflect
the costs associated with imbalances. no adverse effect on competition wiJJ
exist, even 	if 	the 	penalties are -large· in some I bsolute scnse. 	Second, in
some 	situations 	the 	revenues 	from 	these 	penalties 	will 	be 	included 	in 	the
calcula tion of I 	pipeline's allowed rate-of-return. 	In such cases. the firm's
maximum transport rate will presumably be reduced, thus offsetting (It least
partially) the cost-raisin I effect of the penalties.
The 	Minnesota 	Energy 	Issues 	Intervention 	Office 	(EllO) 	provides
another example of possible anticompetitive behavior by pipeline$. 	Although
pipelines cannot charge independent shippers transport rates in excess of the
regula ted 	tariff, 	the 	EllO 	claims 	that 	the 	pipelines 	can 	incorpora te 	other
anticompetitive 	terms 	into 	the 	transport 	aareement 	(such 	as 	those 	dealing
with 	contract duration) that 	have a similar effect: they rlise the 	costs of
11 	Petition of Hadson Gas Systems, Inc., DkL No. RM86·19-o0, August
7, 1986.
18 	An imbalancing penalty is imposed on I 	shipper if its lIS deliveries
~b 	not 	equal 	its 	nominations 	to 	the 	pipeline, 	or 	if 	its 	deliveries 	to 	the
pipeline do not equal deliveries by the pipeline (Hadson submission, p.•6).
10

the independcnt shippers. 	Thc hither cost of ·its rivals allows thc pipelinc's
marketinl affiliatc to raisc its pricc, and earn supracompctitivc returns.
3. PlpeliDe Behulor That MiCht Not be A.tleo.petlthe
Thc 	third 	party 	petitioncrs 	havc 	thus 	idcntificd 	several 	possible
examplcs of Icnuine &nticompctitive bcha vior involving markcti~g affiliatcs.
Howcvcr, thcy also complain of scvcral additional pipeline practices that are
in 	fact 	either 	procompetitive, or 	if 	anticompetitive, then 	unrelated 	to 	the
existcnce of unregulatcd 	marketing affiliates. 	For example, the MinDcsota
ElIO 	claims 	that 	man~ pipelines 	have 	reserves of ·old 	las: that are 	still
subject to FIRe wellhcad price rcgulations. 	The regulated price of this ·old
las· 	is 	below 	the 	current 	market 	price 	of 	las. 	The 	EnO 	contends 	that
pipelines will sell this las to their marketing affiliates, who will then rcsell
it at 	the 	unregulatcd 	price. 	By doing so, tbey increase their total 	returns
carned on the sale of las.
We 	agree 	that 	pipelines 	have 	an 	incentive 	to 	pursue 	this 	strategy.
HowevCf, this behavior is, at worst, without competitive implications; at best,
it is procompetitive. 	The profit margin earned on the resale of ·old las· at
an 	unregulated 	price 	is 	a 	competitive 	rcnt 	that 	is 	attributable 	to 	the
increases 	in 	the 	market 	price 	of 	las 	that 	hive 	occurred 	ove~ time. 	By
internally transferring gas to an unregulated marketing affiliate, In owner of
old 	las 	can 	keep 	this 	rent 	for 	himself, 	rather 	than 	transferring 	it 	to
consumers or 	to another 	broker. 	If las output is unchanged, this practice
will 	affcct 	only 	the 	distribution 	of 	social 	surplus 	between 	producers.
consumers, 	and 	independent 	brokers. 	However, 	it 	seems 	likely 	that 	this
t'
beha vior 	will 	result in 	more ·old las· 	being 	brought 	to tbe market sooner
I )

thaD it otherwise would have been. 	As a aeDeral matter, output expaDsioDs
of this type arc thou&ht to be procompetitive.
The EIIO also Dotes that affiliated marketers have detailed kDowledae
of 	the 	operatioDal 	characteristics 	of 	their 	pipeliDes. 	This 	lives 	them 	an
advantaac in devisiDa optimal 	transportatioD 	routes. 	Althoulh 	we 	discuss
this 	point 	iD 	Ireater detail 	iD section III.B 	below, we 	Dote 	here our basic
view that this alleaed -abuse of information- by marketinl affiliates may be
a lenuine vertical efficiency tbat benefits las consumers.
Similarly, the 	EIIO discusses a situation 	where a pipeline deliberately
underestimates 	its 	-throughput- 	in 	order 	to 	lain 	approval 	for 	a
supracompetitive 	transport 	rate. 	The EIIO maintains that this situation is
especially profitable for those pipelines bavina marketina affiliates, since the
pipeline can charge the affiliate the (lower) competitive transport rate, while
charaing all independent shippers the supracompetitive rate. 	This allegedly
lives the former an advantage when competina for las Illes eontracts. 	We
argue 	somewhat 	differently. 	From 	a 	theoretical 	perspective, 	while 	the
establishment of a supracompetitive tariff constitutes a competitive problem,
it is a problem unrelated to the existence of marketing affiliates. 	Under the
scenario 	described 	by 	the 	EIIO, 	any 	pipeline 	that 	successfully 	obtained 	a
supracompetitive tariff, wbether integrated into marketing or not. would earn
monopoly returns; the existence of aD uDreaulated subsidiary merely transfers
(aDd does not increase) profits within a particular firm.
..
12

111. 	FIRC'. PoUc)' Alter.athe.
Based 	on 	tbe 	CHinas 	of 	tbe 	petitionen 	in 	tbe 	proceedinp discussed
above, and FERC's analysis of tbe competitive attributes of pipeline markets
in 	Order 	No. 	~36, we 	believe 	that 	pipelines 	may 	bave 	market 	power 	in
certain leolrapbie markets. 	Tbus, provided tbat FERC can obtain sufficient
evidence 	tbat 	tbis 	anticompetitive 	bebavior 	is 	actually 	occurrinl 	on 	a
ailnificant scale. there may be an economic justification for FERC to adopt
new policies with respect to tbe currently unrelulued marketiDa affiliates.
In tbis sectioD we discuss the policy options tbat arc available to FERC, IDd
attempt to assess tbe cosu and 	benefits tbat would be associated with each
approach.
A. FERC Should lDcouraat Maximum PlpeliDt CompetltloD
As 	a 	aeDeral 	matter, 	the 	ability 	of 	a 	pipeline 20 	to 	impose
anticompetitive 	terms of sale OD eitber its customers or its input suppliers
rests on a lack of alternatives. 	These competitive alternatives could consist
of other independent pipeline enterprises, or a vertically inteanted pipeline
subsidiary. 	The Justice Department Meraer Guidelines discussion of vertical
meraers is instructive on this point. 	If the structure of the productioD aDd
marketing 	stages of 	the 	natural 	las market 	is otherwise competitive,'1 	the
existeDce 	of 	fully-intearated 	pipeline-marketing 	enterprises 	caD 	preseDt 	a
competitive problem only if (1) produeers have few alternative lnDsportatioD
sources or customers for the las, or (2) there are few alternative sources of
.true.
20 	Or a .roup of colluding firms.
21 	FERC Order No. "36. p. II-4, indicates that FERC believes this to be
13

·as 	(or 	alternative 	fuel 	sources) 	(or 	buyen.· 	It 	must 	Iiso 	be 	costly 	ror
producen 	and 	consumers 	to 	build 	their 	own 	transportation 	capacity.n
Absent 	these 	conditions, 	anticompetitive 	behavior 	by 	intelrated 	pipelines
would be difficult to lustain. 	An attempt to impose a lupracompetitive rate
(or other 	cost-raisinl conditions) on 	a shipper 	would 	obviously rail 	if the
shipper 	could 	Iwitch 	to 	another 	pipeline. or 	build 	in 	own 	transportation
capacity. 	Similarly, a pipeline would not have any market power as a buyer
ot wellhead las if las producers could costlessly switch to other buyers.
Reasoning along similar lines. selective discounts, and other superficially
·discriminatory· 	price 	cuts 	and 	differentials. 	are 	,enerally 	procompetitive
when 	undertaken 	in 	a 	competitive 	environment. 	As 	long 	as 	competition
exists 	among 	pipelines. 	(which 	by 	assumption 	rules 	out 	the 	imposition 	of
anticompetitive 	terms 	upon 	the 	pipeline's 	customers). 	an 	integrated
pipeline/marketer 	will have an incentive to set the 	price u 	of the pipeline
services 	at 	the 	competitive 	level. 	It cannot 	increase 	its 	profits 	by 	doing
otherwise.'· 	Therefore. 	the 	discretionary discounting of 	rates is 	likely 	to
occur 	only 	in 	response 	to 	,enuine 	chanles 	in 	cost 	or 	demand 	when 	the
22 	See 	"Merger 	Guidelines 	issued 	by 	Justice 	Dept.. 	June 	14. 	1984:
Antitrust and Trad, Rtp/ation Rtport Spuia/ Supp/tmtnt. No. 1169, June 	14,
1984. As the Guidelines note. differences in minimum efficient ~le between
the 	different 	stages 	of 	production 	(e.,.. 	between 	.as 	producers 	Ind
transportation) can sometimes make fully-integrated entry difficult. 	If there
are 	larle 	disparities between 	the 	minimum efficient scale of two adjacent
stages 	of 	production. 	In 	attempt 	by 	a 	previously 	nonintegrated 	firm 	to
achieve full integration can be costIy, since it may force the firm to operate
either 	(l) 	at 	an 	inefficiently 	biah 	level 	of 	output 	It 	one 	stage (c.i., 	the
production 	stage). 	or 	(2) 	at 	an 	inefficiently 	low 	level 	of 	production 	at
another slage (e.g., the transportation stage). 	_
S! 	The 	pipeline 	will 	cbarge 	independent 	shippers. 	as 	well 	as 	its
m~rketing affiliate. the competitive transport rate .
. ,. 	For aD analysis of this proposition, sec 	R. Blair and D. Kaserman
(1983), Law and Economics of Vtrtica! Inltlration and Control, pp. 147·51.

pipeline is lubject to competition. 	Because tJ)ese price chlnles are a melns
by which the market adlpu to new condition~ and Ire not dcvica by which
rivals 	are 	hlndicapped, 	they 	arc 	procompetitivc, 	and 	Ihould 	uerefore 	be
cncourlged.
The implication 	of this analysis is that FERC should, 	to the ,rei test
possible extent. adopt policies that increasc thc dClrcc of compc:.ition among
pipelines. 	FERC 	should, 	for 	examplc, attempt 	to 	prcvcnt 	pipelincs 	from
usinl FERC's rcgulatory apparatus as I melns for delaying or im~ding cntry
by 	rivals." 	In 	situations 	whcrc 	I 	pipeline 	is -considering 	adding 	new
capacity to serve customcrs Ilready served by a rival, the incua:~ent has In
incentive 	to 	petition 	FERC 	to 	conduct 	I 	lenlthy 	and 	costly 	=rtification
hearing. 	Tbe 	cost that I 	potential cntrant must incur in respcnse to such
proceedings reduces the expected profitlbility of cntry, and the~:fore makcs
entry less likely. 	The ability of incumbents to deter entry in this manner is
increased 	in 	tbose 	(common) 	situations 	where 	cntry 	into 	tbe 	-threatcned-
market 	is 	a 	minor 	part 	of 	I 	much 	larger 	expansion plan. 	The 	potcntial
entrant 	might 	be 	unwilling 	to 	bave 	its 	entire 	project 	delaye: 	simply 	to
permit entry into one plrticular market. 	As a result, there ma:· 	be a large
number 	of 	instanccs 	where 	otherwise 	profitable 	cntry 	is 	never 	cven
con templa ted.
FERC should 	therefore 	recognize 	that 	incumbents oftcn 	-ill 	have 	In
Inticompetitive 	motivation 	for 	opposing entry. 	Such objcctior.s should 	be
21 	A possible example of this behavior might be found it FERC Dkt.
No. CP86·S74-000 (November 	10, 1986). 	In this matter, Internl:ional Paper
Co. (a pipeline customer) Illeges that Arkla Energy Resources <In interstate
pir>eline) has lodged 	baseless complaints with FERC to prevent the Natural
aas Pipeline Co. from 	building • 	pipeline that would compete with Arkla '5
existing pipeline for International's business.
15

viewed with Ikepticism. and should not (exccpt· in rarc. cascs) be allowed to
delay 	cntry. 	Unless 	thc 	petitioncr 	can 	prescnt 	compellinl 	evidcncc 	that
cntry will 	damaae compctition (and Itot limply a compctitor), we uric that
FERC 	disregard 	the 	protests 	of 	incumbcnts, 	and 	instead 	expedite 	the
certinca tion process. 2e
We 	believe 	that 	competition 	will 	be 	increased 	if 	FERe 	adopts 	the
preceding recommcndation. 	However, we nonetheless recolnize that cven if
FERC is able to obtain the maximum dearee of competition consistcnt with
existing cost and 	demand 	conditions, it 	is still 	possible 	that some 	pipeline
markets will remain susceptible to anticompetitive behavior. 	The following
are FERC's policy alternativcs.
B. R~lul.tory Alt~fD.thes
Throu gh 	the 	creation of 	marketing affiliates, in tersta te 	pipelines ma y
have 	the 	ability 	to circumvent 	reaulation and 	exercise 	both 	monopoly and
monopsony 	power, 	provided 	that 	they 	operate 	in 	markets 	where 	there 	is
little 	competition 	from 	other 	pipelines 	or 	from 	other 	fue!~ 	In 	these
circumstances 	(aDd 	only 	in 	these 	circumstances), 	it 	may 	be 	possible 	to
improve consumer welfare 	throuah 	the imposition of reaulatory constraints
on 	these 	affiliates. 	It 	must 	be 	stressed, 	however, that 	the 	imposition 	of
constraints 	on 	firms 	operating 	in 	competitive 	markets 	is 	both 	unnecessary
and 	inefficient. 	Imposing 	reforms on 	a 	national 	basis. 	without 	relard 	to
2e 	It is theoretically possible that entry deterrence coul~ be efficient
if a pipeline has a nonsustainable natural monopoly. 	In that case. entry by
rivals could raise total production costs and be wasteful. 	While instances in
.... hich such restraint is warranted are likely to be rare, they cannot be ruled
~Qt as 	a 	possibility. 	See 	Baumol, Panzar, and 	Willig 	Conum2blt 	Markets
and 	the 	Theory of Industry Structure, 	(New York: Harcourt 	Brace 	1982), p.
19'7.
16

conditions 	in 	individual 	pipelinc 	·markets,· will 	,cneratc 	costs, 	but 	no
compensatory 	benefits, in 	areas 	where competition 	is vilorous. 	This 	ract
must 	be 	w(ilhed 	carefully 	in 	makinl 	any 	decision 	to 	alter 	the 	custing
relulatory environment.
It unregulated marketing affiliates arc the vehicle by which monopoly
pipelines 	carn 	supracompetitive 	returns, 	then 	several 	potential 	regulatory
approaches to the problem exist. 	These are:
(1) 	regulating 	the 	prices 	and 	margins 	received 	by 	pipelines'
marketing affiliates;
(2) 	reQuiring 	that 	the 	pipelines 	undertake 	the 	complete 	vertical
divestiture of their marketing affiliates; or
(3) 	prohibiting 	the .marketing 	firms 	rrom 	using 	their 	affiliated
pipelines.
In the ensuing sections, we briefly discuss the costs and benefits of each of
these alternatives. 	Depending upon the results of further investigation of
the magnitudes of these benefits and costs, FERC may determine that it is
efficient 	to adopt one these 	policies. 	However, it must also be recognized
that 	retention 	of 	the 	status QUO 	(i.e.. allowing 	regulated 	pipelines 	to 	own
unregulated 	marketing 	affiliates) 	represents 	a 	fourth 	alternative. 	If 	the
welfare 	gains 	associated 	with 	the 	policies 	listed 	above 	do 	not 	offset 	the
corresponding 	welfare 	losses 	(a uri bu ta ble 	to 	the 	loss 	of 	In y 	vertica I
efficiencies, 	Ind 	to 	regulation-induced 	resource 	misallocation), 	then 	the
preservation of the existing environment may on balance constitute the best
of these alternatives .
..
17

1.• 'Iul.tlo. or MarketlDa' AUlIl.tes
Althouah 	the 	NGPA 	of 	197821 	deprived 	FERC 	of 	thc 	aathority 	to
rClulate 	most 	-first 	Ales· 	of 	natural 	las," control 	over 	-sales for 	resalc·
(Co&-. sales from pipelines to local distribution companies) arc ttill subject to
FERC scrutiny ander the Wjust and 	reasonable· rate standard embodied 	in
sections 4 and 5 of the Natural Gas Act.~ 	It has been sUl8este~ by various
interveners 	in 	current 	FERC 	proceedinls so 	that 	sales 	by 	the 	pipeline
affiliates should W 	be entitled to Wfirst sales· status, but ratbe: should 	be
re-elassified 	as Wsales for 	resale: 	thereby placing 	the prices 	a~~er FERC's
review. 	The control of these prices will, it is arlued, deprive pipelines of
the ability to earn monopoly prices through the anticompetitive treatment of
unaffiliated shippers.
The 	FTC 	staff 	urles 	that 	FERC 	reject 	this 	proposal. 	FERC 	has
recognized 	in 	its 	recent 	rulemaldngs 	(i.e., 	Orders 	No. 	4CSI 	It~ 	4C36) 	the
desirability of allowing natural las commodity prices to be estae~.ished by the
interplay 	of 	market 	forces. 	FERC 	bas 	stated 	that 	it 	believes 	the 	las
exploration 	and 	production 	industry 	to 	be 	essentially 	compe:itive. 	This
means 	tha t the 	prices lenerated 	by 	this market arc ·correct: iD the 	sense
thlt 	they 	accurately 	reflect 	society's 	marginal 	valuation 	of II! output, 	as
well as the resource costs that are incurred in the production of this output.
21 	15 U.s.C. 717.
" 	Most 	wellhead 	las sales (with the exception of sc.call::! ·old gas·)
would be unregulated ·first sales: as would any sale to an end-::s:r, such as
an industrial company.
~ 	IS U.s.C. 3301-432.
so 	See Motion to Intervene by Champlin 	Petroleum Co. (Champlin) in
Matter of Tenngasco Corp., Okt. No. CI86-168-000, September I I, 1916.
18

It 	Collows 	Crom 	this 	that 	the 	market 	price 	serves 	u 	a 	proper 	,uide 	to
socially optimal production, consumption, and investment decisioll
We recognize that market prices may be -incorrect- in situations where
pipelines 	can 	confer 	market 	power 	on 	their 	affiliated 	marketers.
Nonetheless, we are skeptical that extension of price controls to cover these
sales 	represents 	the 	best 	policy alternative. 	As 	a 	,eneral 	ma:ter, 	it 	is 	a
dirficul~ if not impossible, task for the regulator to determine the optimal
price 	structure 	that 	should 	prevail 	at 	any 	moment 	in 	time. 	Unless 	the
regulator has access to timely information regarding current Ul~ anticipated
demand 	and 	cost 	conditions, 	and 	can 	process 	this 	informatio: 	efficiently,
there 	is 	a 	distinct 	possibility 	that 	the 	resulting 	set of 	prices 	will 	convey
incorrect 	incentives 	and. information 	to 	all 	market 	participa~:s. 	FERC's
experience 	with 	wellhead 	,as 	pricing 	regulations 	serves 	as 	apt 	specific
illustration of these problems. 	The Department of Energy chancterized the
old 	regulated 	price 	structure 	as 	inefficient, 	distortionary, 	illogical, 	and
unnecessarily 	complex.S! 	When 	added 	to 	the 	direct 	rcscurce 	costs"
associated 	with 	detailed 	price 	regulation, 	the 	costs 	,enerat:d 	by 	these
aUocative distortions would likely more than offset whatever beoefits would
be obtained through the implementa tion of this policy.
Sl 	See FERC Order No. 4S I, p. S8.
S2 	For example, the expenditures on the additional staff that would be
required for carrying out this regulation.
J9

1. 	DIY..tlt.,. or Marketla' ArrlUates
If pipelines are carnina mODopoly returns through vertical intearation
into marketinl, an obvious solution to thu 	u 	simply 	to prohibit rClulated
pipelines 	from 	operatins 	marketins affiliates. 	The 	purpose of this policy
would be to prevent the type of behavior that is aUeaed to have occurred in
the 	lPAMS 	ca~e, 	described 	earlier 	in 	section 	II.B. 	To 	recount 	briefly,
IPAMS, 	a 	,roup of 	producers, alleged 	that PEPL 	and 	CIG (t.... o pipelines)
manipulated the timing of their capacity availability and tariff announcements
in 	a 	manner 	that 	inevitably resulted 	in 	the 	marketing affiliates 	acquiring
rishts 	to 	all 	of 	this space. 	This ,ave the 	marketing affiliates monopsony
power with respect to IPAMS, and (perhaps) monopoly power with respect to
downstream customers.ss
There arc certain clear benefits associated with a divestiture policy. 	It
would 	have low administrative eosu, it would avoid the welfare losses tha t
would 	be 	generated 	by 	a 	distortionary 	regulated 	pricing 	stru:ture, and 	it
would 	doubtless 	deny 	pipelines any 	further 	opportunity 	to exploit 	market
power through the establishment of unregulated subsidiaries. 	The major risk
of 	divestiture 	is 	that 	it 	would 	eliminate 	whatever 	efficien:ies 	might 	be
associated with the intesration of transportation and marketing.
One 	of 	the 	possible 	benefits 	of 	permitting 	pipelines 	to 	retain
unresulued marketins affiliates derives from the larae number of ,as sources
to which the typical pipeline has access. 	The pipeline maybe particularly
well informed about various operational aspects of these fields.. For example,
it 	may 	be 	knowledseable 	about 	the differences 	in the qualit)· of 	the 	las,
SS 	We say ·perhaps,· since we do not have detailed 	kno.... ledae of the
competitive alternatives that might have been available to these customers.
20

K&Sonal variation in demands for eacb tield's output. 'anticipat~ cbanle1in
ntes of flow from eacb field, cxpected exbaustion dates, etc:. 	As a result.
there 	may 	be 	economies 	in 	information 	latherins. 	Tbese 	wormational
economies 	arluably combine 	witb 	the 	economies of aas collection co make
pipelines highly efficient lIS marketers. 	For cxample. tbese two (actors may
endow marketing affiliates witb tbe capability to desian biahly c~:icient sales
COD tracts 	that 	are 	tailored 	to 	tbe 	specific 	requirements 	of 	iDdividual
customers. 	It may Dot be possible for iDdepeDdent marketers te emulate this
activity perfectly. 	Tberefore, tbe opportunity to fashion such C::ltracts may
be lost if pipelines are barred from actina as las marketers.
Tbe 	ability 	of 	marketing 	affiliates 	to 	capitalize 	on 	the:: 	potentially
superior 	access 	to 	info!mation 	miaht 	be 	characterized 	as 	an 	·unfair
advantage- 	by 	the 	independent 	rivals of 	the 	marketing affilia::s. 	Indeed,
some of the petitioners who requested the issuance of tbe curre:t NOI have
araued 	that 	.ceess 	to this 	information 	is a 	principal source c: affiliates'
competitive advantage. 	The Minnesota ElIO, for example, has :laimed that
because the .{filiate -knows the available capacity levels with:: I 	pipeline,
the changes which occur in those capacity levels, and the existe::e of actual
capacity 	'bottlenecks: 	[it] 	can 	determine 	a 	viable 	transpor:ation 	route
acceptable to the pipeline. 	Competitors without this 'inside' info:mation may
not be able to do so.-$4
It is 	quite plausible 	that 	marketing affiliates could capitalize on 	this
type of knowledge. 	It would not, however, be correct to cha:-aeterize such
beha v ior 	as 	an ticompetitive. 	The 	abili ty 	to 	identify 	low 	cost 	transport
roptcs 	would 	appear 	to 	be 	I 	lenuine 	vertical 	efficiency. 	Furthermore,
,.
S4 	Minnesota EIIO peti tion. pp. 10-1 J.
21

becausc of rClulatory constraints, the cxistenct of anunrclulued marketi'nl
affiliate 	may 	increase 	an 	integrated 	pipeline's 	incentive 	to 	exploit 	its
extensive 	knowledle of opportunities for 	minimi%inl the cost of providinl
dclivered las to any particular customer. 	A marketinl affiliate may be able
to 	retain 	(1$ profit) at 	leut part of the cost sa vinls from desilninl 	more
efficient 	delivery 	arranlemenu 	(e.I., 	by 	arranlinl 	11$ 	SWlpS 	with 	other
pipelines, or rearranging its deliveries to other customers), wbile a relula ted
pipeline (which confronts regulations on 	both its las Ind 	transport prices)
would 	eventually 	be compelled 	to reduce its delivered 	price 	by an amount
equal to these savings. 	A third party may not have sufficient information to
replicate 	this 	function 	independently," 	and 	cannot 	(because 	of 	relularory
constraints) 	provide 	financial 	incentives 	for 	the 	pipeline 	to 	make 	these
arrangements. 	Even 	if 	I 	third 	party 	knew 	of 	such 	opportunities, 	it 	is
questionable whether pipelines would have incentives to become ·open access·
carriers under the Order No. ~36 program if they were not permitted to take
advantage 	of 	these 	opportunities 	themselves 	(i.e.. 	through 	their 	marketing
affiliates). 	But 	unless 	pipelines elect to 	join the 	Order No.•36 program,
there 	may 	be 	little 	pipeline 	capacity 	available 	for 	sale 	to 	independent
shippers. 	Clearly, without Iccess to pipeline space, third parties would find
it difficult to exploit these opportunities.
We 	lack 	sufficient 	information 	to 	identify 	and 	measure 	all 	of 	the
efficiencies that might derive from a vertically intelrated market structure.
However, we find it plausible that such efficiencies may exist. 	FERC may be
able 	to 	determine 	wbether 	such 	hypothesized 	efficiencies 	are 	present 	by
.. 	n 	The 	Minnesota 	EIIO 	submission 	(pp. 	10-11) 	clearIL.•vueststhat
third 	parties do not 	have Iccess to the 	information necessary to carry out
tbis function. 	But see footnote 36.
22

analY1ini 	the 	behavior 	of 	marketinl 	affiliates' 	market 	sharn 	Pipelines
would 	not 	be able 	to 	make 	anticompetitive use 	of 	marketin& 	affiliates 	if
competition exists in pipeline markets. 	If one observes larac market shares
for 	marketing 	affiliates 	in 	competitive 	markeu, 	tben 	the 	existence 	of
vertical 	efficiencies 	would 	be 	clearly 	sUllested. 	The 	opposite 	conclusion
would hold, however, if large shares arc observed o1lly in those environments
where the affiliated pipelines possess market power.
If 	the 	evidence 	sUllests 	the 	presence 	of 	efficiencies. 	then 	the
implementation of a divestiture policy could be justified on economic arounds
only 	if 	the 	associated 	wdrare 	aains 	(attributable 	to 	the 	elimination 	of
anticompetitive behavior by pipelines) would more than offset the value of
these 	foregone 	vertical 	efficiencies. 	We do 	Dot 	kDOW 	if 	these 	conditions
would hold. 	There may, however, be alternatives to divestiture that achieve
the same welfare gains, at a lower social cost.
3. 	ProblbltlDI the MarketlD& Firms From UIIDC Tbeir Arriliated
PipeliDes.
A 	Jess 	restrictive 	alternative 	to 	ouuiaht 	divestiture 	is 	to 	deny 	or
restrict marketing affiliates 	from 	booking any capacity on 	their affiliated
pipelines. 	We refer to this as a ·DO access· policy. 	Affiliates would still be
free 	to bid 	for 	sales contracts with 	downstream customers (in competition
with 	other suppliers, such as 	producers and other 	brokers), but they could
Dot contract directly with the affiliated pipeline for transportation. 	Either
the customer 	would 	make 	a 	separate 	transportation 	arrangement 	with 	the
pipeline. or the affiliate could buy space from another independc;nt _lhjpper
~c!.g., another 	broker) 	which 	held 	title 	to space 	on 	the 	affiliated 	pipeline.
and make its own delivery arrangements.
23

Whether 	this 	policy 	reprcscnt, 	a 	ICnsible 	alternative 	to 	divestit~re
depends 	upon 	several 	(actors. 	ODe 	of 	these 	(actors 	is 	the 	aature 	and
ma,nitude of the efficiencies from vertical intearation. 	I( these efficiencies
derive 	principally 	from 	an 	affiliate's 	ability 	to 	capitalize 	on 	its 	detailed
tnowledae 	of 	iu 	parent's 	operational 	characteristics, 	then 	a 	-no 	access·
policy 	could 	entail 	substantial 	welfare 	10sses. Sf 	In 	this 	insunce, 	it 	is
difficult to see how the competitive implications of a no access policy would
differ in any meaningful respect from a policy of full 	vertical divestiture.
Tbis is because one of the attractions to producers or LDCs of dealing with
a 	marketing 	intermediary may be the avoidance of the need to search for
the least cost transportation route. 	They may wish to pay others to perform
this 	service. 	However, 	if 	they 	deal 	with 	marketinl 	affiliatC5 	(who 	are
otherwise 	attractive 	because 	of 	their 	access 	to 	diverse 	sources of 	supply,
etc.), 	under 	a 	·no 	access· 	policy, 	customers 	would 	find 	it 	necessary 	to
undertake 	tasks 	(i.e., 	arranainl 	for 	transportation 	and 	deliver~) that 	they
would 	prefer 	to 	a void. 	In 	some 	instances 	this 	would 	ren:!t 	in 	their
patronizing aD (otherwise) less desirable intermediary, entailinl • consequent
loss 	in 	efficiency. 	Further, 	as 	we 	argued 	earlier, 	the 	cx:hanges 	of
information between pipelines and their affiliates, characterilec by some of
the 	petitioners 	as 	anticompetitive, may actually 	be efficient. 	If so, 	these
efficiencies would be lost under a no access policy.
Sf 	This 	assumes 	that 	independent 	third 	parties 	could 	not 	lenerate
equivalent 	efficiencies. 	As 	we 	araued 	earlier, 	it 	is 	not 	clear 	that 	an
independent third party could ever amass sufficient information to perfectly
4uplicate the functions currently performed by marketing ,ffiliates. 	Other
third 	parties (e.I., see 	the Minnesota EIIO submission, PP. 	10-11) appear to
share this view.
24

It 	may 	be 	the 	case, 	by 	contrast, that 	there 	are 	vertical 	efficiencies
that do not depend upon the exploitation of this knowledae, but wbich would
nevertheless 	be 	sacrificed 	under 	a 	policy 	of 	divestiture. 	. UDder 	these
circumstances, a no access policy would represent a preferred alternative to
divestiture, since it would preserve these efficiencies.
From a 	different perspective, a 	potential drawback 	to the 8 no access·
policy is the 	possibility that it might not prevent anticompetitive behavior.
Even 	if a 	customer 	arranges 	for 	its 	own 	transportation, the 	pipeline 	will
know the source of the aas. 	It may therefore be possible for the pipeline to
impose anticompetitive terms on those shippers who do not buy 115 from the
marketing affiliate. 	If alS customers are aware of this possibility, they will
have 	an 	incentive 	to 	purchase 	gas 	from 	the 	marketing 	affiliat:, 	and the
affiliate's profits will then 	be ·artificially· 	enhanced due to th: pipeline's
exercise of market power.
41. RecommndatloD
The 	paucity 	of 	information 	on 	the 	nature 	Ind 	magnitudes 	of 	the
affiliate-related competitive 	problems Ind 	Iny 	possible vertical efficiencies
necessarily 	makes 	our analysis 	highly 	speculative. s1 	The 	limited 	available
data 	(which 	consists 	mainly 	of 	the 	submissions of third 	party petitioners,
such 	as 	the 	Minnesota 	EIlO) suggests 	that competitive problems may exist.
They 	also suggest 	that 	most of 	the efficiencies potentially associated 	with
the ownership of a marketing affiliate would be foregone under I 	no access
31 	According to FIRe's summary of Tenngasco's Irgumenu, Tenngasco
ha~ claimed that its efficiency IS a marketer is unrelated to iu status IS a
ptpeline 	affiliate. 	See 	FERC 	Order 	Sttlil'lg 	Ctrti/icau 	ApplictUiol'l 	for
Htaril'lg. Tenngasco Corp. Ind Tenngasco Exchange Corp., Dkt. t'o. CI86-168-
000, September 11, 1986, p. 7.
25

policy. 	If 	true. 	this 	would 	suuest 	thlt 	the 	divesture 	policy 	would 	be
optimal. because it eliminates the Inticompetitive problems with certainty. It
• 	low administrltive cost. However. we do not believe that these rubmissions
constitute 	I 	sufficiently 	comprehensive 	body 	of 	information 	to support 	I
definitive recommendation favoring divestiture.
Given 	this 	uncertainty. 	if 	FERC 	finds 	that 	silnificant 	competitive
problems 	exist. 	for 	the 	present. 	the 	-no access· 	policy 	represeDts 	the 	best
alternative. 	This 	policy has several advantages. 	First. it avoids 	the 	risks
associated 	with 	price 	regulation. 	Second. 	in 	contrast 	to 	the 	divestiture
approach. 	it 	provides 	FERC 	with 	I 	considerable 	degree 	·of 	flexibility.
Subsequent to the adoption of this policy. FERC will have the opportunity to
observe a new industry equilibrium. 	If, at this new equilibrium, it appeared
tha t 	pipelines 	were 	engaging 	in 	successful 	anticompetitive 	beba vior, 	the
divestiture remedy could readily be invoked. 	Conversely, if the cost of this
policy 	(i.e., the 	value of 	any 	foregone 	efficiencies) appeared 	te 	be 	larger
than the value of the corresponding benefits, FERC could easily rescind its
order. 	and 	authorize 	the 	marketers 	to 	engage 	in 	direct 	transa:tions 	with
their affilia ted pipelines. 	The divestiture approach. by contrast, is reversible
only at a much higher cost.
IV. CODclusioD
We continue to believe that FERC should adopt policies tha:, as far as
possible. allow competitive forces 	to determine prices and 	outp:.:ts. 	In 	the
present 	inquiry. FERC wishes to determine whether vertical integration 	by
pipelines 	into 	the 	marketing 	(i.e., 	the 	purchase 	and· sale) 	of 	natural 	las
.
allows 	the 	pipelines 	to 	disadvantage 	their 	rivals 	in 	a 	way 	that 	reduces
26

competition. 	If 	so, 	FERC seeks 	,uidance 	in 	selecting 	the 	optimal 	policy
response to this problem.
Based 	on 	available 	information, 	it 	appears 	that 	in 	certain 	markets,
behavior of this sort may be possible. 	However, we note that the evidence
that a competitive problem exists is, at this point, largely anecdotal. 	At the
same time, it appears possible that there are efficiencies aenera ted by these
vertical relationships. 	Therefore, should FERC determine that a competitive
problem exists, the optimal policy will depend upon the relative magnitudes
of 	(l) 	the 	welfare 	aains 	associated 	with 	the 	elimination 	of 	the
anticompetitive 	behavior, 	and 	(2) 	the 	welfare 	losses 	(e.g., 	attributable 	to
foregone 	vertical 	efficiencies, or 	regulation-induced 	resource 	misallocation)
incurred when a particular policy is imposed.
FERC has open 	to it a variety of policy options. 	We believe 	that it
would be extremely unwise for FERC to attempt to reBulate the prices and
margins 	received 	by 	marketing 	affiliates. 	The 	historical 	experience 	with
.... ellhead 	gas 	pricing 	regulation 	suggests 	that 	such 	reaulatioc 	would 	be
inefficient and distortionary. 	Rather, ....e submit that FERC should choose
from the remaining three alternatives.
If 	FERC 	has 	amassed 	sufficient 	information 	to 	indicate 	that 	the
anticompetitive phenomena described 	in section II are likely to occur on a
significant 	scale, 	then 	....e 	believe 	that 	FERC 	should 	adopt 	a e no 	access·
policy. 	The 	principal 	virtue of this strategy lies in 	its flexibility. 	If the
cost of foregone 	vertical efficiencies appeared to be .reater than originally
anticipated, 	FIRC 	could 	remove 	the 	restriction 	and 	permit 	marketers 	to
purchase 	space 	from 	their 	affiliated 	pipelines (i.e., 	restore 	the rtatus quo).
, .
If 	the 	policy 	proved 	to 	be 	ineffective 	as 	a 	means 	for 	deterring
27

anticompctitivc 	bcha vior, 	FIRC 	could 	then 	require 	outrilht 	vcrtical
divestiture. 	Moreovcr, as a tirst step, FIRC should consider suenlthening
its ·open 	access· policy 	by 	requiring 	pipelines 	to 	live 	advaDee.d 	notice 	of
nrm transportation capacity, and by adopting policies allowin, the rilht to
interruptible space to be vested in las purchasers.
28

Appndlx
In 	the 	Notice 	of lDquiry, FERC has requested 	that 	interested 	parties
provide 	answers 	to 	a 	set 	o{ 	specific 	questions 	rClardinl 	the 	allelcdly
anticompetitive behavior o{ pipelines and their marketinl a{Ciliata. 	In this
section, 	we 	respond 	to 	lome 	o{ 	those 	questions, 	or 	indicate 	where 	our
previous analysis discusses those issues.
J. 	{FERC 	question 	• 	J} 	Which, 	il 	ony. 	01 	Ihe 	GbcY'-fM1ttioned
onticomptllllye 	ond 	discrimination 	C01tcer1ts 	can 	be 	reduced 	to
ilUilnilicance 	by 	increased 	compttition 	omonl 	pipelines? 	Which 	ones
will 	not 	be 	omelioraud 	by 	increased 	competition? 	Which 	orus 	will 	be
worsened by competition.'
Response: 	As we discuss in section 	UI.A, above, we believe 	that
all of the concerns discussed in the NOI can be elimiuted, liven
sufficient competition among pipelines. 	However, as -e also note
in 	section 	lILA, 	it 	is 	possible, 	liven 	the 	cost 	atd 	demand
conditions for las transportation, that some pipeline markets could
remain 	susceptible 	to 	anticompetitive 	behavior 	wbe: 	it 	is 	not
possible to increase competition.
2. 	{FERC 	question 	• 	2} 	What 	measures 	01 	compttmor: 	should 	tht
[F ERC J dtvelop 10 	lauge 	the 	in1tnsity 01 competition lor ,as ond lor
transportation in piJrticular markets?
Response: A useful starting point for measuring competition in any
market 	is 	an 	index 	of 	market 	concentration. 	USiIli 	the 	DO]
Merger 	Guidelines 	approach,sa 	the 	·market· 	consists 	of 	the
smallest 	area 	within 	which 	a 	hypothetical 	cartel could impose 	a
·small 	but 	significant 	and 	nontransitory· 	price 	increase. 	The
•• 	sa 	See 	~erger 	Guidelines 	issued 	by 	Justice 	Dept., 	June 	J~, 	J984,"
Antitrust and Trade Regulation Report Spuial Supplement, No. 1169, June 	14,
1984.
29

.0
concentration 	index 	provides some 	information 	about 	the 	likely
stability of such a cartel. 	Tbe luaer the number of firms in the
cartel. 	the 	less .aluable is any 	Cirm's share 	in 	the cartel's joint
profits. 	and 	the 	treater 	is 	tbe 	incentive 	of 	each 	member 	to
undercut tbe cartel price aDd expand output. 	Other factors beld
constant. tbe biaher tbe concentration index. the more likely that
the 	colludinl suppliers can establish and enforce 	tbe output and
price restrictions necessary to maximize joint profits.
One must. bowever. also pay attention to the ·o:her factors·
(e.l.. ease of entry, bomoacneity of products and buyers, extent of
Donprice competition. maanitude of iDterfirm cost differences) that
affect the likeUhood of colJusion in a market. 	Collusion may be
difficult 	to 	sustain, 	despite 	bigh 	concentration, 	if 	other
characteristics 	of 	the 	market 	make 	mutual 	agreement 	untenable.
One 	may 	conclude, 	upon 	evaluation 	of 	these 	items, 	that
competitive outcomes Ire possible, even when tbe market consists
of only I few competitors.
The 	DOJ 	Merger 	Guideliness~ recommend 	the 	use 	of 	the
HerfindahJ-Hirschman Index (HHI) of concentration. 	The HHI is
determined 	by 	the 	distribution 	of 	market 	shares 	(or 	the 	entire
market. 	but 	aives 	proportionately 	areater 	weight 	to 	the 	market
shares 	of 	the 	largest 	firms. 	Under 	specific 	well-defined
theoretical 	conditions. 	the 	HHI 	can 	be 	shown 	to 	be 	related 	to
SQ 	rd.
30

certain 	meuures 	of 	industry 	perfofmance. 40 	Howeve:. 	we 	note
that 	in 	instances 	where 	all 	(irms 	have 	sufficient 	capacity 	to
support 	IUle 	increases in 	output, 	the 	size 	distributio: 	or 	firms
may 	not 	be 	particularly 	important. 	In 	these 	situations, 	lOme
measure of the number of competitors in the market may Jutrice.
An 	HHI 	statistic 	calculated 	solely 	(rom 	info~atiorl on
incumbents' market shares does not tell us whether th: :artel will
be profitable if it were to attempt to collude to reduce output arid
increase 	price. 	Whether 	the 	attempted 	collusion 	is 	profitable
depends, 	to 	a 	larle 	exteDt, 	OD 	the 	size 	of 	potenti1: 	eDtrants.
Studies 	of 	the 	horizontal 	structure 	of 	pipeline 	mark::s 	tend 	to
overestimate 	the 	market 	power 	of 	pipelines 	if 	no 	Ic~ustment 	is
made for potential entry.
The 	identity 	of 	potential 	entrants 	should 	be 	rigorously
defined in the context of antitrust markets. 	Explicit lSsumptions
or 	estimates should 	be 	made 	about 	the 	output 	elas:i:ity 	of 	the
competitive fringe, the elasticity of market deman~. cd 	the size
of 	the 	dominant 	rirm(s). 	The 	selection 	of 	poteI::.:al 	entrants
therefore 	considers 	a 	number 	of 	factors 	that 	arc 	expected 	to
affect the likelihood, of successful collusion.
Nearby 	pipelines should 	be considered 	potentia: entrants 	if
they are large enough 	to expand output sufficieotly :n the local
market 	to undercut 	any collusive 	agrcement 	which 	u:luded 	the
.0 	See 	R. 	Dansby 	and 	R. 	Willig. 	-Industry 	Performar.:e 	Gradient
b~exes: Amtrican Economic Rtvitw, 69 (June 	1979), 249-60, ID~ J. Ordover,
~/ 	al.. 	-Herfindahl 	Conccntration, 	Rivalry, 	and 	Mergers: - HCno:rd 	LAw
R~)·it ....·. 95 (1982),1857-74.
31

potcntial 	entrant. 	This 	'rcQuires "that 	thc 	potcntial 	cntrantbe
located sufficiently close to the market to justify economically thc
investmcnt 	in 	the 	pipeline 	hookup. 	In 	addition, 	the 	potC1ltial
cntrant 	must 	have 	access 	to 	sufficicnt 	throulhput 	(t..... 	excess
capacity 	or 	interruptible 	Illes 	during 	peak 	demand 	periods) 	to
expand 	sianificantly 	output 	in 	the 	ncw 	market. 	In 	the 	shorter
run, 	therefore, 	sales data 	may 	represent 	a 	better 	proxy 	Cor 	the
ability to supply new markets than capacity data, if the potential
en tran t has other contractual commitmen 1$.
A 	similar 	analysis 	should 	also 	be 	performed 	on 	current
suppliers. 	It must 	be dctermined whether thcre arc Iny Cactors
that 	would 	co~strain the ability 	of 	thcse 	firms 	to expand 	their
output in the face of a cartel price. 	Thus, a complete Inalysis of
competition 	in 	natural 	las 	markets 	is 	much 	broad:r 	than 	an
examination 	of 	the 	degree 	of 	concentration 	ameng 	current
producers, or of the effects of cntry on concentration.
J. {FERe qu~stio" .. 4] Should {FERC] impose jurisdiction oYer the
mark~li"g affiliatts 01 pi~li"es?
Response: 	FERC 	should 	not 	rcgulate 	the 	prices 	received 	by
..
marketing 	affiliates 	(sec section 	III.B.l). 	Jurisdiction 	should 	be
established only to permit the exccution of the policies described
in section III.B.3.
4. {FERC q~stion .. 6] Should {FERC] adopt a per se ~le tJu;::
(a) 	Marketing 	affiliates 	not 	be 	permitted 	4ccess 	to l1tt
affiliaud pi~li"e.
Response: See sections III.B.2 and III.B.3.
32

$. 	{FERC qwstioll - 	J$) 	SJuJuid G distinction 	be lNUie 	N~ell G
u:le 	by G marut;II, G//iliatt 0/ Gil ;nttrstatt p;~lille wlUcil ueJu
to 	trQltS port 	tM 	,as 	which 	;s 	I1"rclu1Jed 	Gild 	sold 	tlar~11t 	tM
G/fi/iGttd 	pi~1i1lt. 	4S 	opposed 	to 	olher 	trG1IJI2Cti01lJ 	by 	SllCh
GffiJiatts.'
Response: See section III.B.3.
6. {FERC qwslioll - 	2J) What olher lMasures should {FERC; take
to 	pre.,tllt ""d~ discrimi1U1.tion 	by 	illtuslau 	pipelines 	in 	f",or 	01
lheir marketi1l1 G/filiatts throulh the Iraming of preferences.'
Response: 	If 	the 	·open 	access· 	policy 	is 	to 	succeed 	U1 	markets
where 	conditions 	indicate 	that 	a 	competitive 	problc:. 	is 	likely,
FERe 	must 	modify 	its 	procedures 	to 	make 	it 	easier 	for
independent 	shippers 	to 	obtain 	both 	firm 	and 	interruptible
transportation 	space. 	Currently, 	pipelines 	are 	DOt 	required 	to
live 	advance 	Dotice 	of 	the 	availability 	of 	capacity 	(and 	the
corresponding tariff) for firm transportation. 	Independ:nt shippers
cannot arrange for space if they are not aware of its existence or
its 	price. 	This 	failure 	to 	disclose 	iDformation 	ha3 	allegedly
facilitated 	aDticompetitive 	behavior. 41 	Some 	shi~pers 	ha ve
contended 	that 	pipelines 	have 	refused 	to 	inform 	tb:m 	of 	the
Ivailability of space Ind rates for particular routes ut:il Ifter all
of this space has been booked by the pipeline's markcti.::g affiliate.
In some markets this behavior may leave many buyen and sellers
of las ...ith few competitive alternatives to the marketi.cg affiliate.
Pipelines typically have substantial advance kno... :edge of the
impending availability of capacity for 	firm trlnsport1tion. 	It is
.' 41 	See 	Independent 	Producers 	Association 	of 	Mountain 	Stites 	v.
Panhandle Eastern Pipeline Company and Colorado Interstate Gt3 Company,
FERC Dkt. No. CP86-584-oo0, June 20, 1986.
33

·.
possible, 	therefore, 	that 	cOmpetitioD 	in 	these 	markets 	could 	be
iDcrcued 	if' pipeliDcs were required to Dotify the 	public of the
existeDce of this space (as well as the accompaDyiDI raw) well in
advaDce of its actual bookiDa. 	If such a policy were adopted in
markets 	susceptible 	to 	anticompetitive 	eUecu, 	iJadependent
Ihippers 	milht. 	uDder 	the 	-first 	come, 	first 	lerve~- capacity
allocatioD rule cstablisbed iD Order No. 436, hive U 	opportuDity
to reserve tbe space that is Decessary for the executioc of certain
types of sales contracu. 	FERC may therefore wish te perform I
more 	detailed 	analysis of 	this 	proposal, to determinc whether 	it
would lencrate beDefits iD excess of Iny offsettiDa cosu.
In 	the 	casc 	of 	interruptible 	transportation. 	F"ERC 	must
develop I specific policy tor determiniDI queue positions tbat does
Dot discouraae competitioD amon I 	las suppliers. 	It appears tha t
under 	FERC's current policy, an 	end-user whicb 	is DOW 	buyiDI
(rom a marketiDI affiliate may fiDd itself -lent to thc eDd of the
line- 	(or 	interruptible space 	if it switches 	to aD 	alternative 	las
supplier. 	Tbis 	loss 	of 	space 	will 	be 	costly 	to 	cnd-users;
accordinlly, it lives them aD incentive Dot to ,witeh to competing
las suppliers.
34

·.
.. .
These 	inceDtives 	could 	be 	altered 	by 	developiDI 	I 	policy
whereby 	the 	rilht 	to 	iDterruptible 	apace 	could 	be 	ycsted 	in
purchasers. rather thlD suppliers. 	This would provide buyen with
I 	.reater 	delfee 	or 	flexibility 	iD 	cboosiDI 	amoDI 	competing
sources 	or 	liS, 	Ind 	thus 	increase 	the overall 	competitiveness of
naturll las markets. 	We SUllest that FERC evaluate the beDefits
Ind costs that are likely to be ,eDerated by such a policy.