Phoenix%20Center%20Conference

1
Phoenix Center Conference

• Robert Willig
• October 1, 2004

2
Agenda (i)

• In these times of telecom revolution, weve got
  to get it right.
• Last mile is still bottleneck maybe not for
  very long, but still.
• Special access shows major monopoly power, and is
  bottleneck too evidence from natural
  experiment.
• Overarching competitive problem of access

3
Agenda (ii)

• Access as competitive problem is not unique, and
  recognition is bipartisan.
• Anticompetitive incentives opportunities.
• Forms of predation are not unlikely here.
• Unbundling _at_ TELRIC good solution to save
  intramodal competition.
• TELRIC concept is the competitive price

4
Agenda (iii)

• Unbundling _at_ TELRIC does not depress investment.
• No inconsistency in policy between saving
  intramodal competition and fostering telecom
  revolution with its possible intermodal
  competition.
• Fostering competition needs fast commitment to
  change in inter-carrier compensation for
  competitive neutrality, reduction of regulatory
  risk and efficiency.

5
Gotta Get It Right

• Set the policy structure so that the right, but
  highly unpredictable, decisions are made by real
  competition in the markets that link consumer
  wannas with technological possibles.
• Dont, as policy officials, aspire to decide on
  what should be user prices, services and supply
  technologies.
• Particularly crucial when, as now, whats at
  stake includes where telecom is going, what roles
  will it play in society, and what substantial
  facilities investments will be made.

6
Still Last Mile Bottleneck?

• How ready will public be to switch from wireline
  voice to VOIP in sufficient to control loop
  pricing in 2 years? 4 years?
• How ready will public be to switch from wireline
  voice to cellular in sufficient to control loop
  pricing in 2 years? 4 years?
• How independent of ILEC local services will be
  non-ILEC affiliated cellular carriers if they are
  a principal source of competition?
• What are the odds that deregulation will be in
  the pubic interest in 2 years?
• Lets get policy right for either unpredictable
  possibility!

7
Monopoly Power Over Special Access

• Structural analysis shares and entry barriers
  make market power clear for most of market.
  ILECs have 90 --- Scale economies and sunk
  costs here make substantial entry barriers.
  Absolute cost disadvantages substantial too.
• Natural experiment of FCC Pricing Flexibility
  Order Phase II rate freedom for about 50 MSAs
  where triggers are met (based on wire centers
  with colocation, as proxy for SA entry
  possibilities) no rate declines and 10 35
  continuing rate increases relative to where still
  caps. (over 5 billion overall) Meanwhile, book
  and forward looking costs steadily down since
  1995. SA rates 200 -- 400 of TELRIC and way
  above historic costs. Substantial market power
  revealed!
• Special access is bottleneck for business
  services over 90 of buildings served by IXCs
  have only ILEC fiber access -- entry barriers.

8
Why Public Utility Regulation?The Three Leg Test

• 1. Natural Monopoly in relevant market
• two options cost too much
• 2. High barriers to entry
• sunk costs and scale economies together deter any
  entry
• 3. Sufficient demand for significant monopoly
  profit or rent

• Competition cannot work
• Private supply best nevertheless
• Regulate to mimic competition

9
Bottleneck
BI
A
a
Competition stage
Pe
Pi
DI
De
V
ci
ce
Consumer market
Telecom Electricity Railroads Gas and Oil
Pipelines
CATV Airport Rights CRS IP
ATMs Copy Machine Parts OS and Applics Pubs and
Beer
Rockets and Satellites Main Frame
Parts Physicians and Labs Automobile Radios
10
Anticompetitive Vertical Practices

• Bottleneck is opportunity
• Incentives? versus one-lump theory.
• Regulation gives incentives inadvertently
• Price discrimination
• Diminution of competition in non-coincident
  markets
• Removal of competitive constraint on bottleneck

11
Predation

• Use of bottleneck to enlarge monopoly power is a
  form of predation, or exclusionary vertical
  conduct. Not garden variety price predation.
• Hallmarks are market power upstream, causation of
  increase in market power by making competition
  from active and potential rivals ineffective and
  intent as revealed by conduct.
• Sacrifice test would conduct make sense for
  bottleneck holder but for the harm to
  competition?
• Vertical anticompetitive price squeeze can well
  be such a form of predation. (eg access price
  above downstream retail price, or leaving too
  little pricing room for efficient rival)

12
Policies to Consider

• Vertical separation Bell, Electrics,
  pipelines, MS? Intuit, Rocket-Sat., Merger---
• Change Regulatory Mechanism US RRs prices
  CAPs and freezes
• Mandate Equal Access ONA, unbundling
• AT WHAT PRICES??
• Cost-based Regulation of Access Prices US
  Telecom Policy Act of 1996
• Reactive Regulation of Access Prices Antitrust
  litigation US RRs WHAT STANDARDS??

13
Bottleneck Access Prices for Efficient and Fair
Competition

• should permit efficient suppliers to prevail
  in the market while discouraging inefficient
  supply
• incumbents and entrants should pay the same
  price for the same bottleneck service (if costs
  are unequal then parity of mark-ups)

14
Revenues from Bottom-Up Economic Costs Approach
May Be Less Than Regulatory Rev. Requirements
DUE TO

• Common and fixed costs
• regulatory burdens
• inadequate depreciation
• inefficiencies in operations
• overearnings
• imprudent investments
• cross-subsidies from regulated to non-regulated
  activities
• ____________________________________________
• Competitively neutral (externalized) solutions or
  
• Top-Down Approach

15
Cost Concepts for Competitive Prices

• Long run costs like TELRIC are forward
  looking from today, cost minimizing, and
  unconstrained by the firms past investment
  decisions.
• Short run (and medium run) costs are still
  cost-minimizing looking forward from today, but
  unlike long run costs reflect a planning period
  in which investments in long-lived assets
  inherited from the past remain sunk.

16

• Short run costs include all forward-looking new
  expenditures needed during the planning period.
• They also might include capital costs on the
  inherited sunk assets themselves.
• Economists have identified three alternative
  approaches to the valuation of the sunk assets
  with corresponding costs.

17
One Approach No Value for Sunk Assets with No
Opportunity Costs

• Since the investment is sunk, there is no
  opportunity cost of using it. (That is, if you
  decide not to use the assets, no costs of their
  financing are thereby saved or avoided likewise,
  if you decide to use them, there are no
  additional such costs that result.)
• Short run costs, so defined, are necessarily
  equal to or less than long run costs.
• This follows because in the short run scenario,
  one way to produce the same outputs would be to
  ignore the sunk assets and buy all inputs fresh.
  Hence, if the owner of the sunk assets elects to
  continue using them (as incumbent carriers
  typically do), doing so must be as cheap as, or
  cheaper than, starting fresh.

18
The Reproduction Cost Approach

• The reproduction cost approach values sunk assets
  according to book with physical-life based
  depreciation.
• If the enterprise unconstrained by any sunk
  assets would efficiently choose exactly the same
  assets that had been chosen historically, then
  SRC LRC.
• If intervening technological progress and other
  changes in economic circumstances would change
  those choices, then the appraised value of those
  sunk assets is less than their reproduction cost.
  So a measure of short run cost that includes
  their reproduction cost is systematically biased
  upward.
• That is why this approach violates economic
  logic.

19
Economic Valuation of Sunk Assets

• The economic approach would assign costs to the
  inherited sunk assets according to their
  appraised value.
• The appraised value of the assets is the savings
  their use would permit an enterprise in the
  business, as compared to not using the assets and
  starting fresh.
• If the sunk assets were not used at all, the
  enterprise would incur long run costs.
• Since the appraised value makes the enterprise
  indifferent between using the sunk assets or not,
  this standard leaves the enterprise with SRC
  LRC.
• So TELRIC yields calculated costs that are the
  short run or real costs based on economic
  valuation of sunk assets!!!!

20
example

• Book cost of old sunk assets is 1.0/year
• Forward costs of using the old with efficient
  additions is .5/year
• Costs of same services from new efficient
  facilities is 1.25/year.
• Appraiser would revalue old sunk assets to have
  cost of capital of .75.
• Competitive price is 1.25/year and smart
  supplier uses old assets this way.

21
This Profound Fact Has Been Known Since at Least
1970!!!

• "If the economic value were correctly stated on
  the books the addition of gross return on that
  net book value to the variable costs of operating
  the old plant would produce a cost of service
  exactly equal to that of a new plant." 
•       Alfred Kahn, Economics of Regulation, page
  121 (1970). 

22
Pricing Based on such Costs is Compensatory

• Current prices are based on current efficient
  costs with efficient long-run choices of assets.
• The expected decline from original value to
  current value of sunk assets is included in
  economic depreciation, and thus in prices.
• The chances that actual declines in value will
  deviate from the expected are risk factors in the
  cost of capital.
• This is all so in competitive pricing and in
  efficient pricing based on economic costs.

23
TELRIC Dominates

• TELRIC is a superior regulatory approach
• Minimizes misincentives to over spend and
  cross-subsidze
• Historic cost and real cost methods produce
  sharp misincentives that are costly for consumers
  and anticompetitive.

24
Telecommunications Act of 1996 and
Infrastructure Investment Empirical Evidence
Robert Willig Princeton University
William Lehr MIT
John Bigelow Princeton Economics Group
Stephen Levinson ATT (formerly)
25
Does TA96 unbundling reduce ILEC incentives to
invest?

• Investment Deterrence Hypothesis
• TA96 unbundling reduces ILEC investment
• Denies ILEC opportunity for fair return on
  investment
• Encourages CLEC free-riding on ILEC
  infrastructure

• Competitive Stimulus Hypothesis
• TA96 unbundling enable foster CLEC
  competition
• Competition drives innovation, lowers prices,
• and expands markets
• ILEC and CLEC investment increases

26
Key Issue -- Since TA96 Unbundling So Important
to Competition

• Local facilities still largely natural
  monopolies.
• ILECs still protected by substantial entry
  barriers
• Cable still far from competitive in telephony and
  business broadband .. DBS also.
• So unbundling only way to foster competition in
  services that utilize local plant, and in
  subsequent facilities deployment.
• Access to UNEs at TELRIC prices could be salient
  force for movements to competitively efficient
  pricing, along with competitively neutral
  measures for public purposes.

27
United States Telecom Association, et al. v.
Federal Communications Commission and United
States of America
There are plainly two sides to the effects on
investment of ubiquitously available UNEs at
Commission-mandated prices. . . . The question is
how such investment compares with what would have
occurred in the absence of the prospect of
unbundling, . . ., an issue on which the record
appears silent. Although we cant expect the
Commission to offer a precise assessment of
disincentive effects (a lack of multiple
regression analyses is not ipso facto arbitrary
and capricious) we can expect at least some
confrontation of the issue and some effort to
make reasonable trade-offs. . . .
28
Economic Logic Supports the Competitive Stimulus
Hypothesis Over the Investment Deterrence
Hypothesis

• Under TELRIC principles, UNE rates give ILECs
  adequate incentives to invest since they cover
  risks and economic depreciation.
• CLECs paying such compensatory UNE rates are not
  free-riders whose anticipation deters
  investment by the ILECs.

29

• UNEs can allow a CLEC to overcome entry barriers
  to build a customer base and then transition to
  its own facilities.
• CLECs have strong reasons to invest in their own
  facilities to avoid dependence on their rival
  ILECs once they have the scale.
• Such competitive threats give ILECs added
  incentive to improve their networks in order to
  avoid losing customers to new entrants.

30
Principal Empirical Questions

• What is the relationship between pricing of
  UNEs and investment in network infrastructure by
  ILECs?
• Investment Deterrence Hypothesis Positive
• High UNE prices discourage utilization by
  CLECs.
• Less utilization by CLECs encourages ILEC
  investment.
• Competitive Stimulus Hypothesis Negative
• High UNE prices discourage entry by CLECs.
• Reduced competition attenuates ILEC incentives
  to invest.
• ANSWER The estimated relationship is negative.

31
More Principal Empirical Questions

• Do the data support the mechanism of the
  competitive stimulus hypothesis?
• Is there a negative relationship so that
  lower UNE prices encourage CLEC activity?
• Is there a positive relationship between CLEC
  activity as a driver of ILEC investment?
• ANSWER YES YES -- Together these effects
  show how lower UNE prices stimulate ILEC
  investment.

32
Structural Form Equations
33
Reduced Form Equation
34
Data

• Investment Data
• ILECs FCC ARMIS reporting system
• State by state
• Largest ILECs (BOCs)
• CLECs Generally not available
• Many CLECs privately held.
• Many are part of larger entities, and
  investment in telecom network infrastructure is
  not consistently reported in sufficiently
  disaggregated form.
• Measures of CLEC activity are available
• Number of firms active by state.
• Counts of Zip Codes within states with CLEC
  service.

35
(No Transcript)
36

• Exhibit 1 shows ILEC investment has negative and
  statistically significant relationship with UNE
  price.
• Reduced form relationship accounts for over 77
  of state to state variation in ILEC investment.
• ILEC investment increases with Labor Force share
  in FIRE, Population Growth, Average Revenue
• ILEC investment decreases with TELRIC

37

• Exhibit 2 shows Relationship between number of
  CLECs and UNE price is negative and statistically
  significant. Relationship between ILEC investment
  and number of CLECs is positive and statistically
  significant.
• Approximately 75 of variation in ILEC investment
  explained, and 45 of variation in number of
  CLECs.
• Both relationships are statistically
  significant overall at a high confidence level.
• ILEC investment increases with Population
  Growth and Average Revenue, and decreases with
  TELRIC
• Relationship between number of CLECs and TSR
  Discount is positive and statistically
  significant at 90.

38

• So policies to assure competitive access to
  bottlenecks at competitive prices are
• helpful, not dangerous to investment
• likely necessary for consumer protection.
• What do we need to prepare for fuller
  competition?
• Reform intercarrier compensation

39
Intercarrier Compensation

• Contributions to universal service funding must
  be competitively neutral.
• What should be equivalence scale???
• Opportunities for receipt of universal service
  support must be competitively neutral -- who is
  an ETC? targeting?
• Terminating access fees should be curtailed
  even if monopoly power over local customers
  erodes with intermodal competition, terminating
  access still has much more monopoly power payer
  has less control than called
• Do-not-call list, caller id, call duration,
  calling relationships all move cost causation
  toward called party.
• Most costs at originating and terminating ends
  not volume sensitive. MOVE TO MORE COMPLETE SLC!