FTC Staff Comment Before the Postal Rate Commission Concerning Mail Classification

V900016
COMMISSION AUTHORIIED
BEFORE THE
UNITED STATES OF AMERICA POSTAL RATE COMMISSION
WASHINGTON, 	D.C. 	20268-001
Procedures 	for 	Consideration 	of
Contract 	Rules, 	1989-1990
COMMENTS OF
Docket 	No. 	RM89-5
THE STAFF OF THE 	BUREAU OF ECONOMICS 	OF THE
FEDERAL TRADE COHMISSION1
(submitted February 	20, 	1990)
1 These 	comments 	represent 	the 	views 	of 	the 	staff 	of 	the
Bureau 	of 	Economics 	of 	the 	Federal 	Trade 	Commission. 	They 	are
not 	necessarily 	the 	views 	of 	the 	Commission or 	any 	individual
Commissioner. 	Questions 	about 	these 	comments 	may be 	addressed 	to
John C. 	Hilke, 	Federal 	Trade 	Commission, 	Bureau 	of 	Economics, 	6th
Street 	and 	Pennsylvania Avenue, 	N.W., 	Washington, 	D.C. 	20580,
telephone: 	(202) 	326-3483.

BEFORE THE
UNITED STATES OF AMERICA POSTAL RATE 	COMMISSION
WASHINGTON, 	D.C. 	20268-001
Procedures 	for 	Consideration of
Contract 	Rules, 	1989-1990 	Docket 	No. 	RM89-5
CODEH'l'S OF
THE STAFF OF THE 	BUREAU OF ECONOMICS 	OF THE
FEDERAL TRADE COMMISSIONl
(submitted February 	20, 	1990)
I. 	INTRODUCTION'
The 	staff 	of 	the 	Bureau 	of 	Economics 	of 	the 	Federal 	Trade
Commission 	(FTC) 	appreciate 	the 	opportunity 	to 	submit 	these
comments 	to 	the 	Postal 	Rate 	Commission 	(PRC) 	concerning 	the
desirability of 	reorienting 	"mail 	classification 	to 	deemphasize
the 	unitary character of 	at 	least 	some 	classes 	or 	subclasses, 	and
treat 	them more 	as 	groupings 	of 	separable 	operational 	functions"
(separable 	functions). 	Of 	the 	several 	questions 	posed by 	the 	PRC
1 These 	comments 	represent 	the 	views 	of 	the 	staff 	of 	the
Bureau 	of 	Economics 	of 	the 	Federal 	Trade 	Commission. 	They 	are
not 	necessarily 	the 	views 	of 	the 	Commission or 	any 	individual
Commissioner. 	Questions 	about 	these 	comments 	may 	be 	addressed 	to
John 	C. 	Hilke, 	Federal 	Trade 	Commission, 	Bureau 	of 	Economics, 	6th
Street 	and 	Pennsylvania Avenue, 	N.W., 	Washington, 	D.C. 	20580,
telephone: 	(202) 	326-3483.
1

in 	this 	matter, 	we 	limit 	our 	comments 	to 	those 	concerning
economic 	costs 	and benefits: 	questions 	1 	and 	3. 2
The 	mail 	classification 	system of 	United 	States 	Postal
Service 	(USPS) 	usually 	bundles 	a 	variety of 	potentially 	separable
postal 	functions. 	Functions 	within 	the 	bundle 	include 	collection
of 	mail 	through a 	local 	network of 	pick-up 	points, 	sorting of
mail, 	transportation of 	mail 	from 	one 	local 	area 	to 	another, 	and
distribution 	of 	mail 	through a 	local 	network of 	delivery 	routes.)
Unbundling 	refers 	to 	the 	policy of 	providing one 	or more 	postal
functions, 	separate 	from 	the 	traditional 	bundle 	of 	services, 	at 	a
separate 	price.
Our 	analysis 	indicates 	that 	the 	economic 	desirability of
offering unbundled postal 	services 	(in 	addition 	to 	bundled
services) 	depends 	on 	(1) 	the 	prospective 	cost 	savings 	from
unbundling, 	(2) 	the 	effects 	of 	unbundling on 	the 	true 	costs 	of
services 	that 	remain offered exclusively by 	the 	USPS,4 	and 	(3)
2 There 	are 	other concerns 	(e.g., 	the 	intent 	of 	the 	Postal
Reorganization Act, 	etc.) 	that 	the 	PRe may wish 	to 	consider 	in
its 	decisionmaking 	in 	this 	matter. 	We 	limit 	our comments 	to 	the
likely effects 	of 	various 	unbundling possibilities 	on economic
efficiency and consumer welfare.
) This 	list 	is 	an 	illustration of 	how postal 	functions 	might
be 	separated. 	This 	is 	not 	intended 	to 	be 	a 	definitive 	list 	of
separable 	postal 	functions. 	We 	assume 	in 	the 	comment 	that most,
if 	not 	all, 	mail 	must 	at 	least 	go 	through 	the 	collection,
sorting, 	and 	delivery 	functions, 	whether or 	not 	the 	USPS 	performs
all 	these 	functions.
4 By 	"true cost" 	we mean 	the 	opportunity costs .of 	the
resources 	used. 	When we 	discuss 	cost-shifting by 	a 	regulated
monopolist 	we 	are 	discussing accounting costs 	that 	are 	allocated
to 	a 	function. 	These 	costs 	may 	not 	reflect 	opportunity 	costs 	and
may be 	fictitious.
2

the 	effects 	of 	unbundling on 	the 	incentives 	and 	ability 	of 	the
USPS 	to 	shift 	costs 	from 	the 	competitive 	functions 	to 	the
remaining 	regulated 	functions 	or 	segments.
If 	unbundling a 	function 	is 	likely to 	provide 	cost 	savings
to 	consumers 	and 	it 	is 	unlikely 	to 	increase 	the 	costs 	of 	the
remaining 	services 	of 	the 	USPS, 	then unbundling 	is 	likely 	to 	be
economically desirable 	due 	to 	increased competition among 	postal
service 	providers.
If 	unbundling 	is 	unlikely 	to 	provide 	current 	cost 	savings,
but 	it 	is 	also 	unlikely 	to 	increase 	the 	true 	costs 	of 	other
services 	of 	the 	USPS, 	unbundling may 	still 	be 	desirable 	because
it 	may encourage 	the 	USPS 	to 	adopt 	future 	cost-saving 	techniques.
At 	the 	same 	time, 	society will 	bear 	little 	cost 	from 	unbundling
under 	these 	circumstance 	(since 	consumers 	have 	no 	incentive 	to
patronize 	higher cost 	suppliers).
If 	unbundling 	is 	likely 	to 	increase 	the 	true 	costs 	of 	other
USPS 	services, 	whether or not 	it 	is 	likely 	to 	decrease 	costs 	of
the 	particular 	funct~on that 	is 	unbundled, 	then unbundling may
still 	be 	desirable. 	In 	this 	case, 	imposing 	a 	fee 	on 	private
~uppliers of 	this 	function, 	to 	offset 	the 	increased 	true 	costs 	of
the 	USPS, 	may be 	an 	attractive 	option. 	Such a 	fee 	could 	preserve
incentives 	to minimize 	costs 	as 	well 	as 	avoid 	unwarranted
increases 	in 	the 	costs 	of 	the 	USPS. 	The 	fee 	should be 	equal 	to
the 	increase 	in 	the 	costs 	of 	services 	provided exclusively by 	the
USPS.
3

Regardless 	of 	whether 	unbundling 	lowers 	or 	raises 	the 	true
costs 	of 	other 	functions, 	however, 	the 	PRC 	will 	need 	to 	consider
whether 	allowing competition 	in some 	functions 	might 	allow 	the
regulated 	firm, 	USPS, 	to 	inefficiently 	"shift costs" 	from
competitive 	functions 	to 	their 	regulated monopoly 	functions 	(if
any 	remain). 	Such cost-shifting could, 	if 	left 	unchecked, 	thwart
any 	potential 	entry 	and 	result 	in 	higher overall 	postal 	service
pricing.
II. 	FEDERAL TRADE COMMISSION STAFF EXPERTISE
The 	FTC 	is 	an 	independent 	regulatory 	agency 	responsible 	for
maintaining competition and 	safeguarding 	the 	interests 	o~
consumers. 5 	The 	staff 	of 	the 	FTC, 	upon 	request 	by 	federal,
state, 	and 	local 	government 	bodies, 	often 	analyze 	regulatory 	or
legislative proposals 	that 	may 	affect 	competition or 	the
efficiency of 	the 	economy. 	In 	the 	course 	of 	this 	work, 	as 	well
as 	in 	antitrust 	and consumer 	protection research 	and 	litigation,
the 	staff 	apply established and 	recent 	developments 	in economic
theory 	and empirical 	analyses 	to 	competition and consumer
protection policy 	issues, 	including efficiency rationales 	for
rate 	regulation 	and entry 	restrictions.
The 	FTC 	Bureau of 	Economics 	staff 	have 	commented 	previously
on various 	issues 	before 	the 	PRC. 	Of 	particular 	relevance 	to
this 	inquiry 	are 	our 	comments 	on 	the 	Complaint 	from 	the 	Third
5 	15 	U.S.C. 	Section 	41 	et 	seq.
4

Class 	Mail 	Association 	(PRC 	Docket 	No. 	C89-1) 	and 	our 	comments 	in
the 	Monopoly Theory 	Inquiry 	(PRC 	Docket 	No. 	RM89-4).6
III. 	COSTS AND BENEFITS OF UNBUNDLING POSTAL SERVICES
(Question 	1)
A. 	Introduction
In considering 	the 	advisability 	of 	unbundling 	separable
postal 	functions, 	the 	PRC 	may wish 	to 	compare 	the 	total 	costs 	of
providing 	postal 	services 	to 	consumers 	under 	the 	current 	system
and 	under 	an 	alternative 	system 	in which mail 	classes 	and 	rates
are 	unbundled. 7
Under current mail 	service 	classifications 	and 	the 	private
express 	statutes,8 consumers 	of 	some 	classes 	of 	mail 	generally
6 These 	comments 	were 	filed 	on February 	28, 	1989 	and
September 	I, 	1989, 	respectively.
7 The 	term 	"consumers" 	is 	used 	here 	to 	denote 	ultimate
consumers 	of 	postal 	services 	as 	well 	as 	intermediate 	buyers 	of
postal 	services 	whose 	product-prices 	charged to 	ultimate
consumers 	reflect mailing costs.
8 Our comments 	do 	not 	address 	the 	potential 	conflict 	between
the 	private 	express 	statutes 	and 	full 	unbundling 	of 	postal
functions. 	Full 	unbundling, 	i.e. 	consumers 	can buy 	all, 	some, 	or
none 	of 	their 	postal 	services 	from 	suppliers 	other than 	the 	USPS,
should be 	efficient 	if 	no 	postal 	functions 	exhibit 	the
characteristics 	discussed 	in 	Section 	III.C. 	If 	some 	functions
have 	these 	characteristics, 	unbundling 	the 	rest 	of 	the 	postal
functions 	should still 	maximize 	benefits 	and minimize 	costs 	so
long 	as 	significant 	cost-shifting 	from 	competitive 	functions 	to
the 	remaining 	regulated monopoly 	functions 	does 	not 	occur. 	For
the 	services 	with 	the 	characteristics 	discussed"in 	Section
5

must 	purchase 	an 	entire 	bundle 	of 	services 	from 	the 	USPS 	or 	do
without 	these 	postal 	services 	entirely.9 	In contrast, 	a 	buyer of
postal 	services 	in 	an 	unbundled 	system would 	be 	likely 	to 	face 	a
wider 	array of 	choices 	similar 	to 	that 	faced 	by consumers 	of
other 	goods 	and 	services. 	For many products 	and 	services, 	the
consumer 	has 	at 	least 	three 	choices 	in 	addition 	to 	buying 	a
complete 	bundle of 	services 	from 	a 	single 	supplier. 	These
choices 	include: 	producing all 	the 	services 	oneself, 	producing
some 	services 	and 	buying 	some, 	and buying all 	the 	services 	from 	a
variety of 	suppliers.
In 	some 	cases 	the 	cost 	of 	buying 	a 	bundle 	of 	services 	from 	a
single 	producer 	is 	much 	less 	than buying the 	same 	services
separately. 	In 	such cases, 	most 	cons~ers will 	naturally elect
to 	buy 	the 	bundle 	of 	services, 	but 	no 	restriction on 	unbundling
the 	service 	is 	necessary 	to 	obtain this 	cost minimizing 	result.
For 	example, 	most 	consumers 	elect 	to 	buy 	their 	automobiles
III.C., 	a 	system 	of 	fees 	(discussed 	in Section 	111.0.), 	rather
than 	entry restrictions, 	may provide 	the most 	efficient 	results.
Unbundling 	is 	likely 	to 	result 	in one 	or more 	suppliers
offering a 	wide 	variety of 	combinations 	of 	postal 	functions
ranging 	from 	single 	functions 	to 	combinations 	similar to 	those
currently offered by 	the 	USPS. 	Actual 	cost 	functions 	and
competitive 	incentives 	then would determine 	which combinations
are 	sufficiently popular to 	be 	offered.
9 Some 	limited unbundling options 	have 	been made 	available
in 	the 	form 	of 	discounts 	granted when a 	mailer 	performs 	a
particular task, 	thereby 	relieving 	the 	postal 	service 	of 	the
costs 	associated with 	this 	task. 	Sorting discounts 	and 	nine-
digit 	zip 	code. discounts 	are 	examples. 	Distance 	related 	rate
differences 	are 	also 	a 	form 	of 	unbundling 	because 	consumers 	have
the 	option of 	obtaining 	a 	lower 	rate 	by 	transporting 	their 	items
(in 	bulk) 	closer to 	the 	final 	destination before 	entering 	them
into 	the 	USPS 	mail 	stream.
6

already 	assembled 	because 	the 	costs 	of 	gathering 	parts 	and
separately contracting 	for 	a~sembly are 	relatively 	high. 	No
restriction 	on 	such 	contracting 	is 	necessary 	to 	make 	bundling 	of
automobile 	assembly 	functions 	almost 	the 	universal 	choice 	of
consumers.
In other 	instances, 	some 	consumers 	may 	be 	able 	to 	produce
one 	or more 	services 	in 	a 	bundle 	at 	lower cost. 	In 	such
instances, 	required bundling of 	these 	services 	may 	result 	in
higher 	costs 	and 	less 	satisfaction 	for 	consumers. 	For 	example,
many consumers 	continue 	to 	prefer to 	prepare 	and cook 	their 	own
food 	rather 	than 	to 	buy prepared meals. 	Unbundling 	in 	this
instance 	allows 	some 	consumers 	to 	assemble 	and cook 	their own
food 	while 	other consumers 	can 	purchase 	prepared 	items 	or 	dine 	in
restaurants. 	Regulations 	which 	require 	consumers 	to 	purchase 	all
food 	items 	in prepared 	form 	would 	increase 	costs 	to 	society and
force 	many consumers 	to 	pay prices 	above 	their own costs 	and 	to
obtain 	services 	they do 	not 	want.
When consumers 	in 	a 	market 	economy 	face 	a 	"make 	or 	buy"
decision about 	an 	activity 	(separable 	function), 	they 	have 	a
profit 	incentive 	to 	buy 	that 	service 	from 	the 	lowest 	cost
supplier(s), 	and 	society 	also 	benefits 	from 	this. 	In 	some
narrowly defined cases 	(discussed below), 	however, 	the cost
savings 	from 	unbundling particular 	services 	may 	result 	in 	higher
costs 	of 	other 	services 	provided by 	the 	USPS 	that 	off-set 	the
former 	gains.
7

We 	now 	turn 	to 	the 	questions 	of 	incentives 	to 	be 	the 	lowest
cost 	suppliers 	(Section B) 	and 	off-setting cost 	increases
(Sections 	C and 	E).
B. 	Differences 	in 	Incentives 	to 	Minimize 	Costs
From 	an economic 	perspective, 	consumers 	are 	generally 	better
off 	under 	a 	competitive market mechanism, 	unless 	there 	are
substantial 	market 	failures 	that 	can only be 	mitigated by
restricting buyer choices 	to 	a 	monopoly 	supplier. 	Competition
provides 	incentives 	to 	improve 	economic 	performance, 	that 	is, 	to
minimize 	costs, 	to 	improve 	quality, 	and 	to 	develop 	new 	products
and methods 	of 	production. 10 Without 	competition, 	the
incentives 	to 	minimize 	costs 	may be 	attenuated. 	Unbundling would
increase 	consumers' 	choices 	of 	suppliers 	and 	therefore 	increase
competition 	in postal 	services.
Under 	a 	mail 	classification system with unbundling, 	if 	the
USPS 	offers 	the 	lowest 	rate 	on 	any particular 	function 	or 	set 	of
separable 	functions 	and 	its 	service 	is 	of 	equivalent 	quality,
consumers 	can be 	expected to 	patronize 	the 	USPS 	for 	those
functions. 	And, 	if 	the 	USPS 	is 	the 	lowest 	cost 	producer 	of 	all
10 This 	is 	the 	fundamental 	economic 	rationale 	for
competitive markets. 	See, 	for 	example, 	Scherer, 	F., 	Industrial
Market 	Structure 	and Economic 	Performance, 	2nd edition, 	Chicago:
Rand McNally, 	1980.
Incentives 	to 	innovate 	do 	not 	guarantee 	that 	specific
innovations 	will 	be 	successful 	or 	that 	the 	timing 	of 	innovation
will 	be 	optimal. 	Innovation 	typically 	involves 	a 	significant
risk 	of 	failure.
8

separable 	and unbundled 	functions 	and 	sets 	prices 	close 	to 	its
minimum costs, 	the 	USPS 	will 	remain 	the 	monopoly 	supplier. 	If
the 	USPS 	is 	not 	the 	lowest 	cost 	producer, 	however, 	an 	unbundling
of 	supply 	could 	give 	consumers 	an 	increased opportunity 	to 	reduce
their 	postal 	costs 	by patronizing 	lower 	cost 	suppliers 	or 	by
supplying the 	functions 	themselves.
In our Monopoly Theory Comment 	(Docket 	RM89-4), 	we 	discussed
theoretical 	reasons 	for 	concern 	that 	production under 	a 	protected
monopoly 	need 	not 	minimize 	costs. 	Recent 	extensions 	of 	economic
theory 	in 	the 	area 	of 	"agency" 	relationships, 	including 	the
theories 	of 	bureaucracy, 	x-inefficiency, 	and 	public 	choice,
indicate 	that 	government monopolies 	(unconstrained by
competition) 	are 	likely 	to 	be 	inefficient. 	Competition 	through
the 	unbundling 	of 	services 	could 	help 	offset 	this 	inefficiency 	by
providing 	incentives 	for 	the 	USPS 	to 	minimize 	costs.
In 	the 	theory 	of 	bureaucracy, 	competitive 	pressure 	from
customers 	who 	provide 	their own 	services 	and 	from 	other 	providers
can establish 	limits 	on 	the 	costs 	of 	the 	government 	sector 	by
providing elected officials 	with 	increased 	information on minimum
costs 	and on 	alternative 	levels 	of 	service. 	With better 	informed
officials, 	managers' 	effort~ to 	inflate 	an 	enterprise's 	budget
will 	increase 	the 	probability 	that 	the 	government 	enterprise will
be 	displaced by outside producers 	(private 	or 	public) 	or
otherwise more 	strictly constrained. 	In 	the 	x-inefficiency
theory, 	the 	possibility of 	outside 	production makes 	excessive
costs 	more 	obvious 	and 	likely 	to 	draw 	attention. 	This 	may
9

constrain cost 	increases 	and may 	improve 	group 	norms 	of 	work
effort. 	In 	public 	choice 	theory, 	alternative 	suppliers 	of 	an
initially monopolized 	service 	reduce 	the 	relative 	organizing
advantages 	of 	the 	managers 	of 	the monopoly 	by 	forming 	well-
organized and well-informed countervailing 	interest 	groups 	that
often oppose 	the 	interests 	of 	budget-maximizing government
enterprises 	or challenge 	coalitions 	between politicians 	and
enterprise managers 	that 	exclude 	competition. 11
C. 	The 	Problem of 	Increased Costs 	of 	Other 	Postal 	Services
Allowing consumers 	to 	select 	the 	lowest 	cost 	suppliers 	for 	a
particular separable 	postal 	function 	may 	not 	be 	optimal 	if 	doing
so 	increases 	the 	true 	costs 	of 	other 	services 	provided by 	the
USPS 	monopoly. 	In particular, 	the 	entry of 	an 	additional
suppljer may 	not 	be 	efficient 	if 	postal 	services 	are
characterized by 	subadditivity.12 	Even 	if 	costs 	are 	subadditive,
a 	natural 	monopoly, 	the 	least 	cost 	industry 	structure, 	may 	not 	be
able 	to 	sustain 	itself 	against 	new entry.13 	If 	a 	natural
11 Also 	see 	Shapiro, 	C., 	and R. 	Willig, 	"Privatization to
Limit 	Public-Sector Discretion," 	unpublished, 	1989.
12 See 	our comments 	in 	the Monopoly Theory 	Inquiry 	(Docket
No. 	RM89-4).
Costs 	are 	subadditive when 	the 	cost 	of 	producing 	some 	set 	of
products 	or 	functions 	is 	lower when 	it 	is 	produced by 	a 	single
firm 	than when the 	same 	set 	is 	produced by more 	than 	one 	firm.
13 	Sustainability means 	that 	the 	market 	demand 	and 	cost
functions 	allow 	the 	monopolist 	to 	price 	such 	that 	(1) 	the 	market
clears 	(i.e. 	the 	monopolist 	produces 	all 	that 	is 	demanded 	at 	that
price), 	(2) 	the 	monopolist 	at 	least 	breaks 	even, 	and 	(3) 	entry 	is
10

monopoly cannot 	sustain 	itself, 	some 	restrictions 	on entry may 	be
economically efficient.
When 	a 	firm 	produces 	several 	products, 	as 	does 	the 	USPS, 	it
may 	be 	especially difficult 	to 	determine 	whether 	the 	firm 	is 	a
natural 	monopoly, 	and 	if 	so, 	whether 	the 	natural 	monopoly 	is
sustainable. 	This 	difficulty 	arises 	because multiproduct
production may make 	costs 	dependent 	upon 	the 	mix 	(scope),14 	as
well 	as 	the 	quantities 	(scale),15 	of 	the 	different 	products
produced.
A full 	discussion 	of 	the 	conditions 	under 	which 	a
multiproduct 	natural 	monopoly exists, 	and when 	it 	is 	sustainable,
is 	available 	in 	the 	literature. 16 The 	chief 	implication of 	this
literature 	can be 	succinctly 	summarized: 	there 	is 	no 	efficiency
basis 	for 	protecting a 	monopolist 	from 	entry 	unless 	there 	is 	a
good 	reason to 	believe 	that 	the 	industry 	is 	an unsustainable
natural 	monopoly. 	Unless 	the 	monopoly 	is 	unsustainable, 	market
forces 	will 	induce 	and 	sustain 	the 	natural 	monopoly 	as 	the 	least-
unprofitable.
When costs 	are 	such 	that 	total 	costs 	in 	an 	industry 	are
minimized when 	there 	is 	a 	single producer, 	the 	industry 	is 	termed
a 	natural 	monopoly.
14 Cest 	relationships 	in which costs 	of 	production 	for 	a
product 	are 	lower when 	the 	product 	is 	produced 	in conjunction
with one 	or more other products 	are 	termed economies 	of 	scope.
15 	Cost 	relationships 	in which 	the 	average 	costs 	of
production are 	lower when more 	of 	the 	product 	is 	produced 	are
termed economies 	of 	scale.
16 	See 	Baumol, 	Panzar, 	and Willig, 	op. 	cit.; 	Bailey, 	E. 	and
A. 	Friedlaender, 	"Market 	Structure 	and Multiproduct 	Industries,"
Journal 	of 	Economic 	Literature, 	20 	(1982), 	pp. 	1024-48; 	and
Sharkey, 	op. 	cit.
11

cost 	industry 	structure. 	Indeed, 	allowing 	free 	entry 	into 	the
provision of 	a 	bundle 	of 	postal 	services 	that 	is 	a 	sustainable
natural 	monopoly will 	have 	no 	impact 	because 	no 	firm 	could
profitably enter 	the 	market, 	assuming 	the 	monopolist 	minimized
costs. 17
Similarly, 	if 	conditions 	for 	sustainable 	natural 	monopoly
prevail, 	permitting unbundling will 	have 	no 	effect 	if 	the
monopolist 	is 	already cost 	minimizing 	and 	pricing 	accordingly.
If 	the 	monopolist 	is 	not 	already cost minimizing, 	however,
,
increasing 	competition 	through 	unbundling will 	increase 	its
incentives 	to do 	so.
Although 	the 	discussion 	above 	raises 	the 	theoretical
possibility that 	bundling postal 	services 	under 	a 	protected
monopoly 	represents 	an 	optimal 	industry 	structure, 	there 	is
little 	empirical 	evidence 	that 	separable 	postal 	functions 	are
characterized by 	subadditivity and 	unsustainable 	natural 	monopoly
characteristics. IS 	Neither 	scholars 	nor 	the 	USPS 	have 	provided
17 Although no 	static 	efficiency 	gains 	might 	be 	expected 	in
this 	case, 	gains 	from 	increased 	incentives 	and 	opportunities 	to
innovate 	in 	the 	future 	might 	accompany unbundling.
IS 	For 	a 	discussion of 	empirical 	tests 	for 	subadditivity and
unsustainable 	natural monopoly 	see 	Eyans, 	D., 	and 	J. 	Heckman,
"Multiproduct Cost 	Function Estimates 	and Natural 	Monopoly Tests
for 	the 	Bell 	System," 	in Evans, 	D., 	Ed., 	Breaking Up 	Bell: 	Essays
in 	Industrial 	Organization and 	Regulation, 	New 	York: 	North
Holland, 	1983; 	Evans 	and Heckman, 	·"A Test 	for 	Subadditivity of
the 	Cost 	Function with 	an Application 	to 	the 	Bell 	System,"
American 	Economic 	Review 	74:4 	(September 	1984), 	pp. 	615-623; 	and
our 	testimony 	in Monopoly Theory 	Inquiry: 	Docket 	No. 	RM89-4, 	pp.
372-377. 	If 	cost, 	output, 	and 	input 	price 	data 	are 	available 	for
the 	various 	services 	provided by 	the 	postal 	service, 	it 	should be
possible 	to 	assess 	whether 	the 	necessary conditions 	for
subadditivity 	are 	met. 	If 	these 	conditions 	are 	not 	satisfied,
12

evidence 	of 	these 	conditions. 	We 	suggest 	that, 	barring evidence
of 	such problems, 	a 	policy of 	unbundling 	is 	likely 	to 	be
economically efficient. 19 	If 	the 	PRC 	were 	to 	proceed with 	a
restructuring of 	the mail 	classification process 	to 	unbundle
postal 	services, 	a 	reasonable 	priority 	list 	of 	particular
services 	to 	be unbundled might 	be 	derived by 	identifying
functions 	that 	are 	already widely produced by consumers 	or 	are
contracted out 	by 	the 	postal 	service 	to 	private 	suppliers.
D. 	An Alternative When 	Free 	Entry May 	Increase 	Costs
If 	unrestricted entry 	into 	a 	separable 	postal 	function 	will
(or might) 	increase 	the 	USPS's 	true 	costs 	in other 	services 	where
the 	USPS 	is 	the 	sole 	supplier, 	one 	option 	is 	simply 	to 	forbid
entry 	in 	that 	particular 	separable 	function. 	This 	option,
the 	sustainability 	issue 	need 	never 	be 	addressed, 	and 	a 	strong
case 	can be made 	for 	allowing 	free 	entry 	into 	the 	provision of
all 	postal 	services. 	If 	they do 	exist, 	then 	entry prohibitions
may be 	appropriate 	if 	unsustainability can also 	be 	shown 	to 	exist
and 	no 	economically appropriate 	fee 	system on other 	suppliers 	can
be 	applied 	to 	take 	into 	account 	these 	scale 	and 	scope 	effects 	of
entry.
Anecdotal 	evidence 	suggests 	that 	the 	question of
subadditivity and unsustainable 	natural 	monopoly 	are most 	likely
to 	be 	relevant 	in 	the 	local 	pick-up 	and 	delivery 	functions. 	See
our comments 	and 	testimony 	in Monopoly Theory 	lnquiry, 	Docket 	No.
RM89-4, 	pp. 	380-390.
19 The 	process 	of 	evaluating unbundling proposals 	should
help 	to 	identify 	the 	core 	of 	postal 	services, 	if 	any, 	where
subadditivity or 	unsustainable 	natural 	monopoly conditions 	are
most 	likely to 	exist. 	This 	may 	help 	to 	clarify where 	an
efficiency 	rationale 	for 	restrictions 	on 	entry, 	such 	as 	the
private 	express 	statutes, 	are most 	persuasive. 	Consequently, 	the
information gathered 	from 	evaluating unbundling 	proposals 	could
be 	useful 	for 	the 	PRC's 	Monopoly Theory 	Inquiry 	(Docket 	RM89-4).
13

however, 	may 	lead 	to 	higher production costs 	because 	the 	benefits
of 	competition are 	completely 	lost. 	(See 	Section 	III. 	B.) 	An
option 	that 	preserves 	competitive 	incentives 	for 	the 	USPS 	to
minimize 	costs, 	but which 	also 	avoids 	an 	inefficient 	loss 	of
economies 	of 	scale 	and 	scope 	at 	the 	USPS,20 	is 	to 	charge 	a 	fee 	to
entrants 	that 	compensates 	the 	USPS 	for 	loss 	of 	scale 	or 	scope
economies. 21 Such a 	fee 	could be 	set 	to 	compensate 	the 	USPS 	for
any 	true 	increase 	in 	the 	cost 	of 	other USPS 	mail 	operations. 22
The 	fee 	system would 	allow 	for 	entry by 	low 	cost 	producers 	and
would, 	therefore, 	allow consumers 	to 	pay 	lower 	rates 	for 	postal
services. 23 	That 	is, 	consumers 	would continue 	to 	buy 	bundled
20 	If 	unbundling 	results 	in 	loss 	of 	economies 	of 	scale 	or
scope 	to 	the 	USPS, 	society will 	bear 	these 	higher 	costs
regardless 	of 	whether 	the 	USPS 	is 	compensated. 	If 	alternative
providers 	are 	able 	to 	profitably 	operate 	as 	well 	as 	compensate
the 	USPS, 	the 	cost 	savings 	of 	the 	private providers, 	however,
must 	exceed 	the 	cost 	increases 	in 	the 	operations 	of 	the 	USPS. 	On
net, 	total 	costs 	would 	have 	to 	decline.
21 	See 	our comments 	on 	the 	Complaint 	of 	the 	Third Class 	Mail
Association, 	Docket 	No. 	C89-1.
22 We 	assume 	here 	that 	the 	fee 	is 	calculated 	to 	compensate
the 	USPS 	for 	increased costs 	and 	that 	the 	USPS 	minimizes 	the
in~reases by 	appropriately 	reallocating 	inputs 	just 	as 	a 	profit
maximizing, 	efficient 	nonsustainable monopoly would do. 	Limiting
the 	fee 	to 	this 	level 	will 	give 	the 	USPS 	incentives 	to make 	the
required 	reallocations 	of 	inputs.
23 	See 	our 	co~ents on 	the 	Complaint 	of 	the 	Third Class 	Mail
Association, 	Docket 	No. 	C89-1. 	In 	that 	comment, 	we 	suggested
that 	the 	USPS 	expand 	its 	current 	array of 	rate 	discounts 	to 	allow
a 	rate 	discount 	for 	a 	particular potentially separable 	function 	-
- 	user-delivery of 	addressed 	third 	class 	mail. 	The 	final 	price
of 	the 	stamp 	(that must 	be 	affixed 	to 	all 	mail 	using postal 	roads
or boxes) 	would 	reflect 	the 	loss 	of 	economies 	of 	scale 	and 	scope
at 	the 	USPS, 	if 	any.
A similar approach 	has 	been 	suggested by a 	consultant 	for
the 	Direct Marketing Association 	(DMA). 	The 	DMA 	proposed 	to 	the
USPS 	that direct 	mail 	firms 	be 	allowed 	to 	"rent" 	the 	use 	of 	mail
14

services 	from 	the 	USPS 	if 	the 	bundle 	of 	services 	is 	cheaper 	than
if 	the 	services 	were 	unbundled and 	purchased 	separately.24
The 	preceding discussion 	of 	economic 	efficiency criteria 	and
efficiency enhancing 	policies 	is 	summarized 	in 	Figure 	1. 	In
Figure 	1, 	unbundling 	is 	the 	preferred policy 	if 	there 	are 	no
subadditivity problems 	or 	if 	subadditivity occurs 	in 	the 	context
of 	a 	sustainable 	natural 	monopoly. 	Unbundling with 	appropriately
calculated 	fees 	on non-USPS 	suppliers 	is 	the 	preferred 	policy 	if
subadditivity occurs 	in 	the 	context 	of 	an 	unsustainable 	natural
monopoly. 	In either case, 	however, 	the 	PRC 	must 	ensure 	that 	the
USPS 	cannot 	"shift" 	costs 	to 	any 	remaining 	regulated monopoly
functions.
E. 	Cost-shifting and Misallocation of 	Costs 	under 	Unbundling
While 	unbundling 	has 	significant 	potential 	to 	improve
consumer welfare, 	it will 	do 	so 	unambiguously only 	if 	the 	PRC 	can
boxes 	from 	the 	USPS 	by 	purchasing and 	affixing 	a 	stamp 	to 	each
piece 	that 	would 	reflect 	the 	USPS's 	loss 	of 	overhead 	(not
necessarily 	lost 	scale 	and 	scope 	economies), 	estimated 	to 	be 	3/4
cents 	per 	item. 	(The 	private 	express 	statutes 	currently give 	the
USPS 	exclusive 	rights 	to 	use 	these 	receptacles 	at 	each 	address.)
All 	the 	collecting, 	sorting,. transportation, 	and 	delivery
functions 	would be 	provided by private 	firms. 	(Letter 	to 	Frank
Heselton, 	Assistant 	Postmaster General, 	Rates 	and Classification,
from 	Robert W. 	Inhofe, 	dated April 	14, 	1988.)
24 The 	public 	interest 	will 	be 	served 	best 	if 	care 	is 	taken
to 	determine 	accurately 	the 	appropriate 	magnitude 	of 	the
"delivery discount." 	Like 	other 	cost 	savings 	calculations
considered by 	the 	PRC, 	this 	may prove 	to 	be 	a 	difficult
determination to 	make 	and one 	that 	requires 	considerable 	time 	to
refine.
15

ENTRY
PROHIBIT
NO
SUBADDITIVITY
UNBUNDLE
(ALLOW FREE ENTRY)
FIGURE 1
SEPARABLE
POSTAL
FUNCTION
8U8TAI NABLE
NATURAL
MONOPOLY
UNBUNDLE
IALLOW ,,, ••• N'"")
~SUBADDITIVITY
U N8U8TAI NABLE
NATURAL
MONOPOLY
IS IT FEASIBLE TO
COMPUTE FEES TO
COMPENSATE FOR
EFFICIENCY LOSSES?
...-. 	......V...E......SZL..-,I 	,...,__1l!2:
UNBUNDLE; IMPOSE
FEES ON NON-USPS
SUPPLIERS

constrain efforts 	by 	the 	regulated 	firm 	(the 	USPS) 	to 	"shift" 	the
accounting costs 	of 	serving the 	unregulated 	segment(s) 	to 	the
remaining 	regulated 	segment. 	The 	unbundling 	strategy 	discussed
here 	may 	give 	the 	USPS 	the 	incentive 	to 	shift 	costs 	away 	from
those 	service 	segments 	where 	they 	face 	new competition 	into 	those
segments 	where 	they 	retain a 	regulated monopoly. 	Such 	actions
could 	allow 	the 	USPS 	to 	misallocate 	costs 	and 	to 	deter 	new 	entry
through 	artificially 	lower costs 	or 	lower 	prices 	in 	the
competitive 	segment. 	The 	losses 	generated by 	such 	activity 	could
be 	recouped 	through 	higher 	prices 	(justified by 	the 	higher
shifted costs) 	in 	the 	regulated monopoly 	service 	segment. 	If
such cost 	shifting was 	extensive 	enough, 	the 	prices 	of 	postal
services 	might 	rise 	under 	an. unbundling 	scheme. B
~ For 	a 	discussion 	of 	the 	incentives 	of 	a 	regulated
monopolist 	to 	shift 	costs 	to 	the 	regulated 	segments, 	see 	Brennan,
T., 	"Cross-Subsidization and 	Discrimination by Regulated
Monopolists," 	Economic 	Analysis 	Group 	Discussion Paper 	87-2,
Department 	of 	Justice, 	March 	19, 	1987 	and 	Braeutigam, 	R. 	and
Panzar, 	J., 	"Diversification 	Incentives 	Under 	'Price-Based' 	and
'Cost-Based' 	Regulation," 	RAND 	Journal 	of 	Economics 	20:3, 	(Autumn
1989), 	pp. 	372-91. 	Braeutigam and 	Panzar 	also 	discuss 	additional
welfare 	reducing 	incentives 	of 	profit-maximizing 	regulated
monopolists 	who 	serve multiple markets 	(e.g., 	inefficient
technology choice 	and capital 	investment, 	overproduction 	in
competitive markets, 	and 	failure 	to 	minimize 	costs). 	They 	argue
that 	many of 	the 	incentive 	problems 	might 	be 	overcome 	through
price 	cap 	regulation 	rather 	than 	the 	more 	traditional 	rate 	of
return 	regulation.
'16

IV • 	APPROACHES 	TO THE UNBUNDLING PROCESS
(Question 	3)
If 	the 	PRC 	elects 	to 	pursue 	a 	policy of 	increased
unbun~ling, it may wish 	to 	consider 	three 	alternative
approaches. 26 In accord with 	the 	previous 	discussion, 	we 	assume
that 	the 	objective 	of 	an unbundling policy 	is 	to 	increase
consumers' 	options 	while 	increasing 	incentives 	to 	minimize 	true
costs.
The 	first 	approach, 	the 	primary 	alternative 	identified 	in
the 	request 	for 	comments, 	is 	to 	establish new mail
classifications 	for 	separable 	postal 	functions. 	Under 	this
system, 	consumers 	would 	have 	the 	option of 	buying 	one 	or more
services 	from 	the 	USPS 	or 	from 	other 	suppliers 	based on 	the
relative 	attractiveness 	(price 	and 	quality) 	of 	their 	services.
Both the 	USPS 	and private 	firms 	would 	be 	free 	to 	offer 	directly
competing 	services 	and bundles 	of 	services. 	This 	system
maximizes 	consumer choices 	and provides 	the 	potential 	for
significant gains 	to 	consumers. 	This 	new 	system could, 	however,
increase 	search 	and 	transactions 	costs 	for 	some 	consumers 	as 	they
consider buying postal 	services 	from 	several 	sources 	for 	the
26 A fourth 	approach, 	allowing contracting between 	the 	USPS
and 	specific 	customers, 	is 	not 	considered here, 	as 	requested 	in
the 	PRC's 	notice.
17

first 	time. 	These 	costs 	may 	initially 	inhibit 	full 	consumer
utilization 	of 	the 	new 	range 	of 	postal 	alternatives. 27
A second 	alternative 	for 	increased unbundling, 	one 	in
limited use 	already, 	is 	worksharing. 	In 	this 	system, 	consumers
buy 	at 	least 	some 	services 	from 	the 	USPS 	and may 	garner 	various
discounts 	by performing 	functions 	for 	the 	USPS. 	While 	expansion
of 	this 	system would 	provide 	net 	benefits 	to 	consumers, 	it 	puts
more 	limits 	on consumer 	choice 	than 	full 	unbundling because 	it
limits 	competition 	to 	a 	subset 	of 	separable'postal 	functions. 	In
addition, 	like 	pure 	unbundling, 	worksharing may 	involve
additional 	search 	and 	transactions 	costs 	for 	consumers 	who 	must
separately 	locate 	and purchase 	additional 	postal 	services 	from
other 	sources 	before 	transferring 	the 	items 	to 	the 	USPS. 28
27 	This 	problem may be 	less 	severe 	after 	the 	transition
period' as 	more 	consumers 	become 	used 	to 	mixing postal 	providers
and 	as 	brokers 	arise 	to 	offer combinations 	of 	services 	from
various 	suppliers.
Individual 	consumers 	and 	firms 	with modest 	or 	irregular
postal 	demand 	are 	likely 	to 	have 	larger 	per 	item 	search 	and
transaction costs. 	This 	may prompt 	these 	consumers 	to 	forego
using 	the 	services 	of 	alternative 	postal 	providers 	until 	the
experience of 	other users 	effectively 	reduces 	search costs.
The 	entry of 	brokers 	offering competing bundles 	of 	services,
may 	itself 	require 	additional 	consumer 	search costs. 	However,
these 	brokers 	have 	strong 	incentives 	to 	provide 	advertising 	and
other 	forms 	of 	truthful 	information 	that 	should help quickly
inform consumers 	of 	the 	attributes 	of 	their 	postal 	service
combinations.
28 Here 	too, 	the 	problem of 	search 	and 	transactions 	costs
may be 	less 	severe 	after 	the 	transition 	period 	as 	consumers
become 	used 	to 	mixing postal 	providers 	and 	as 	brokers 	arise 	to
offer 	particular combinations 	of 	services 	from 	various 	suppliers.
As 	in 	the 	case 	of 	pure 	unbundling, 	individual 	consumers 	and 	firms
with modest 	or 	irregular postal 	demand 	are 	likely to 	have 	large
per 	item 	search and 	transaction costs.
18

A third 	alternative 	is 	to 	permit 	the 	USPS 	and 	interested
private 	firms 	to 	transform 	themselves 	from 	vertically 	integrated
single-source suppliers 	of 	postal 	services 	into 	postal 	service
retailers 	offering 	services 	of 	several 	different 	suppliers. 29
(This 	option would be most 	effective 	as 	an 	addition 	to 	increased
unbundling 	through classification changes 	and works hare
programs.) 	The 	USPS 	could offer consumers 	a 	wide 	variety of
services, 	some 	of 	which 	it 	produces 	internally 	and 	some 	of 	which
others 	produce. 3o Competing postal 	retailers 	could do 	the 	same.
For example, 	the 	USPS 	could offer 	3rd class 	mail 	service 	with
sorting conducted by 	the 	USPS 	or with 	sorting provided by 	a
particular private 	firm. 31 The' retail 	outlets 	of 	parcel 	post
firms, 	convenience 	store 	chains, 	and 	oil 	companies, 	for 	example,
would be 	free 	to 	do 	the 	same.
In comparison to 	classification changes 	alone 	or expanded
works hare 	programs 	alone, 	a 	one-stop 	shopping 	approach would
29 	Some 	private 	parties 	have 	already established 	retail 	mail
service 	firms 	that 	offer 	services 	from 	several 	alternative
providers 	when 	a 	service 	(such 	as 	expedited mail 	and parcel 	post)
has 	been opened to 	competition. 	These 	firms 	may also 	offer
options 	for 	additional 	premium 	services 	(such 	as 	enhanced post
office box 	security) 	that 	exceed the 	quality of 	service
customarily provided by the 	USPS.
Under 	this 	scenario, 	the 	USPS 	and 	other 	postal 	retailers
would offer the 	services 	of 	other 	firms 	at 	a 	mark-up.
Competition between postal 	retailers, 	including 	the 	USPS, 	would
help 	to 	restrict 	retail 	mark-ups.
30 With 	its 	wide 	network of 	sales 	offices 	and widespread
familiarity 	to 	consumers, 	the 	USPS 	might well 	have 	a 	significant
first 	mover advantage 	in 	this 	role.
31 	Similarly, 	a 	brokerage 	firm 	could offer 	document 	mailing
with 	air 	transport 	by a 	private 	air express 	company 	and 	home
delivery by 	the 	USPS.
19

increase 	consumers' 	search costs 	the 	least. 	Under 	this 	approach,
even consumers 	with 	relatively 	high 	search 	and 	transactions 	costs
would 	be 	readily 	able 	to 	compare 	the 	prices 	and 	features 	offered
by different 	suppliers 	of 	different 	separable 	functions 	at
familiar 	USPS 	locations. 	These 	consumers 	would also 	be 	able 	to
readily arrange 	to 	obtain combinations 	of 	services 	that 	best meet
their 	price 	and 	quality preferences. 	Under 	this 	alternative,
enhanced competition 	should exist 	between 	the 	USPS 	and 	other
postal 	retailers 	and 	should provide 	incentives 	to 	minimize 	costs,
price 	competitively, 	and 	improve 	the 	quality of 	service.
If 	the 	PRC 	chooses 	to 	pursue 	these 	unbundling 	options, 	and
some 	products 	or 	services 	continue 	to 	be 	produced exclusively by
the 	USPS, 	some 	regulatory precautions 	to 	prevent 	cross
subsidization 	from 	the 	monopolized 	function 	to 	other 	functions
may 	be 	efficient 	and promote 	competition.
v. 	CONCLUSIONS
Reorientation of 	the 	mail 	classification system 	toward
unbundling 	separable postal 	functions 	is 	an 	attractive 	economic
proposition 	if 	doing 	so 	will 	not 	(1) 	raise 	the 	true 	costs 	of
other 	services 	produced by 	the 	USPS, 	nor 	(2) 	result 	in
significant cost-shifting 	from 	the 	unregulated competitive
sectors 	toward 	any 	remaining 	regulated monopoly 	sectors.
Unbundling 	has 	the 	potential 	to 	increase 	competitive 	incentives
to 	minimize 	true 	costs 	in 	the 	unbundled 	functions 	and 	to 	provide
20

net 	benefits 	to 	consumers. 	Where 	unbundling could 	increase 	the
costs 	of 	other 	services 	of 	the 	USPS, 	an 	attractive 	policy 	option
may be 	to 	allow 	unbundling, 	but 	to 	charge 	entrants 	a 	fee 	that
compensates 	the 	USPS 	for 	losses 	of 	economies 	of 	scale 	and 	scope.
These 	unbundling options 	would be most 	viable 	if 	the 	PRe 	can
effectively monitor 	any cost-shifting behavior by 	the 	USPS 	to
ensure 	that 	the 	costs 	from 	the 	newly competitive 	sectors 	are 	not
inappropriately 	allocated 	to 	any 	remaining 	regulated monopoly
sectors. 	Unbundling which 	includes 	giving 	the 	USPS 	an 	option 	to
serve 	as 	a 	retailer of 	postal 	services 	produced by 	independent
firms 	as 	well 	as 	internally may 	be 	particularly 	attractive
because 	it 	could reduce 	search and 	transactions 	costs 	for
consumers.
21