Notes

1/9/2006Revision 0.1

Author: Philip Larson

Telecom & Internet Policy Seminar: Class Notes

Telecom & Internet Policy Seminar: Class notes

Table of Contents

1. Lecture 1...................................................................................................................................3

Philip Larson Page 2


1. Lecture 1

1.1 Telecom & Internet Policy Seminar -

Tom Hazlett – Law 617.Questions

1) Can we get examples of previous paper topics? 2) Docomo – first platform for delivering wireless web capabilities. Telecommunications in China.

Copyright in some countries. BitTorrent.

Lecture 1- Telecoms & Internet: A Study in Dynamic Markets- Paper

o Paper is 90% of the grade. 10% class participation. - Submit a one-page outline for approval by the end of 6th week of class. 15-25 page paper. It is due at

the conclusion of final exams. - Other

o Readings are posted on TWEN. o He is not a lawyer, he is an economist.

- Texto Get the additional required text from the syllabus. Who Controls the Internet? Illusions of a

Borderless World. Also get the optional texts. - Slides- Ground to cover

o Legal regimeso Economics of Telecomso Financial realities in information markets – o Effects of technological change – emergence of railroads, telephone networks, new markets

for communications. The markets sometimes look new and innovative. A deep social understanding of the phenomena requires looking beyond trends in the US. Look at what consumers are doing, what manufacturing is doing regarding devices, and also emphasize what financial markets suggest about the way the world is shaping up.

o Academic literature – he has an academic bias. Law reviews, economic journals. o Policy debates – what regulators are talking about, what courts are doing.

- POTS and PANSo Wired

LANs - “last mile” networks – last mile networks connect end users to larger WANs.

ILECs (Bells) – Incumbent local-exchange carriers. These were the LAN operators.

CLECs (entrants) – competitive local exchange carriers. Companies that were not the original incumbents.

Cable TV – this technology is different. Phones use a standard copper wire. Cable TV is a coaxial cable which brings larger capacity because its video product requires higher capacity.

WANs - Long distance/wide area networks – these are connecting the LANs from one to another. With the telephone, this distinction is becoming less visible. The leading carriers are being acquired by the owners of local area networks. There was an artificial distinction between WANs and last mile networks created by regulations.

o Wireless Broadcasting – point to multipoint – a broadcast generally goes from one

transmitter to a wide audience. Radio, television.



Mobile telephony – this is new. Only been important for about 15 years. Technology it is based on is not new. Cellular telephones were first developed in 1946 but not much happened due to regulatory inertia.

Fixed Wireless – use of point to point (sometimes point to multipoint). These are not mobile technologies. Historically, it was important in the disintegration of the old telephone monopoly. The breakdown of AT&T started in the late 50s early 60s when a small company tried to put in links between major cities that would bypass the long-distance monopoly of AT&T. MCI. Microwave ___ Incorporated.

Satellite – this is of course wireless. Satellite is nice for covering a wide area, over some dimensions. Video delivery. Also important for voice in remote areas where fixed wired systems are costly. This may increasingly be a competitive solution in high population areas with low-orbit satellites. What is the problem with geosynchronous orbits? There is latency. It takes a long time to travel the distance. Latency on standard satellites is a problem. If you put the satellites lower in the sky you can reduce the latency. LEOs – low earth orbit satellites. Mobile satellite services will probably become a direct competitor of telephone service. This could cover a continent pretty easily. However, the costs are still such that competitive pricing is difficult. If you could get guaranteed connectivity via satellite anywhere in the country as a subscriber, that might complement standard cellular subscriptions. Importance of satellite in data communications are especially important in remote areas. Bringing broadband to outlying areas.

- Negroponte Switcho Theory: When we were born, TV was over the air, phone calls over wires. When we die, the

reverse will be true. o Problems w/ Theory – satellite TV, VoIP calls aside. You get the idea.

- World Wide Buzzo Internet – o Social networking sites – MySpace.com. o Skype – responsible for 5% of internet traffic in the US today. o IP video – YouTube – tip of the icebergo iPODso Wi-Fi - o Municipal Wi-Fi – there are lots of articles about municipal Wi-Fi. There are no published

subscriber or user data. Is municipal Wi-Fi being used? Why are there all these articles. Philadelphia and San Francisco are doing this.

- World-wide fixed line vs. Mobile Phone Subscriptions – graph in slides. A lot of people did not expect cellular to be such a big deal. McKinsey report in 1980 predicted 1M subscribers in 2000. Gross underestimate. GSM second quarter is 2B subscribers. CDMA is about 350M for September 2006. You get to about 2.4B subscribers world-wide. Fixed line has grown too but much slower. Mobile phones are therefore the killer application of our generation. It is revolutionizing economies and social relationships in Africa, etc. The trend in the US is that fixed line subscriptions are decreasing even though world-wide they are increasing. China is growing rapidly in fixed lines and mobile technology. In 1993 there was basically zero fixed line subscribers. This has shot upward. VoIP can now provide advanced technology for fixed line systems. In business, there is a preference for fixed lines because the voice quality is better. India is going crazy in wireless. It has the lowest prices in the world ($.04/minute). In total numbers, China is almost double the US.

o Why is all this important? This is where people are putting there dollars around the world. Mobile and fixed line subscriptions are incredibly big business.

- Rival Network Platforms – where have we gone over possible rival competitors? o 1934 Communications Act:

Common Carriage - propounded common carriage national policies – previously there were state policies. This attaches certain obligations to people who do business with the public. They have an obligation to do so without discrimination, etc. Common carriage. You can have common carriage without being a common carrier.



Federal express is not a common carrier but follows common carriage principals. Nondiscriminatory prices and terms.

Regulated natural monopoly - Gov’t would insure there would be service everywhere. You can’t have two telephone companies serving a market. There is a natural monopoly and the government would regulate the rates. Making sure the prices were reasonable and just. The idea has been that the regulator establishes price ceilings at the price they would be in competitive markets. You get the efficiency of monopoly without the higher prices. Prices are set just above cost. Issues came up about when competition was feasible. Certain entrants tried to create competition. Regulators then had to decide whether to permit competitors from trying to challenge. Regulators decided that competition was not in the public interest. All these things were put under a single organization (AT&T).

But why leave anything to chance? Competition is not in the “public interest”

Local long distance & equipment – one system (AT&T). o “Hush a phone” (1950s) – they sold physical attachments that

shielded the conversation someone had. They were for the pre-cubicle environment in offices. It would quiet your phone call so you could make one from your desk without interrupting the other people. AT&T objected to the sale saying it would disrupt its business. It said it had the right to determine what devices connected to their system. While this physical attachment might not look like a threat, it is a slippery slope. AT&T convinced the FCC to block the sale of this device and there was an ensuing legal battle. How natural is it that regulators have to stop these extensions. Why can’t AT&T outcompete Hush a Phone if it has economies of scale? The regulators instinct was not to allow competition. The courts allowed it to take place.

Common carriage – regulators interpret this to mean discrimination against competitors. This is a great irony of the 1934 Communications Act.

Monopoly yields rents, cross-subsidies- Six Decades Pass- Irrational Exhuberance

o 12/5/1996 – speech by Chairman of the Federal Reserve Board “Alan Greenspan”. o Was Greenspan right that this was irrational exhuberance or was this simply a wave of

“creative destruction” in which waves of disruptive technologies come to market and bring great opportunities. However, they also create havoc and eliminate great value.

o MSFT, INTEL, CSCO v. S&P500 from 1985-2005. There was an enormous gain for these companies. There was also loss on 3/10/2000 when the tech bubble burst. After the bust, all three companies have losses that are much higher than the market as a whole.

- An Emerging “New Economy – massive amounts of new capital are being attracted into these “new economy” sectors.

o MSFT, CSCO INTEL – most valuable companies in the world at one time. o Browser War 1994-96 – this created the first mass market in Internet access. This happened

in 1994 because 1993 was the year the Internet was commercialized. That was when commercial access was permitted. In a short time, this was made extremely widespread by the creation of not just the browser (which made access easy) but combining the browser with competitive ISPs (internet service providers). What became the largest was AOL. They were extremely aggressive at distributing its browser-ware by carpet bombing the US (250M sign up disks).

AOL, ‘carpet bombs’, 250M sign up diskso Path Dependence and the race to be first

Pets.com, CDNow, Webvan, Peapod, DrKoop.com – thought to be great business ideas. There were too many of them, it was competitive, execution was poor. eBay



survived and was successful because it actually uses the Internet. You can’t do eBay without the Internet. The crash came brutally for many companies.

o WorldCom’s 24 annual growth forecast. – Bernie Ebbers went to prison for defrauding investors.

o Fiber WANs (wide area networks) – global crossing, 360Networks, Qwest, Williams, Level3, Enron, MCI, PSINet.

All of them went bankrupt except Qwest (which came close) and Level3 which did not take on debt. Enron went under for other reasons, but still. Creative destruction. This was quite an experience.

- A Perfect Stormo 8/1995 – Netscape IPO ~$4B – increased 200% in a single day. This was why Alan

Greenspan called this irrational exhuberance. o 2/1996 – Telecommunication Act – eliminating barriers to entry in telecoms. Republican

Congress and Democratic president signed this bill. They reversed the 1934 Communications Act bias towards the natural monopoly. Competition is now welcome by public policy. Did this opening of the market help cause the irrational exhuberance. Did PPT cause the Dot.com bubble?

o 1995-1999 – 26% annualized returns for the S&P 500 and 40% annualized returns for the NASDAQ. There has never been a 5 year period with these kinds of returns. This changes the game when we talk about Telecommunications as a sector. It used to be a protected market and the variance of return was very low. They were guaranteed by regulation and monopolies were guaranteed certain returns for investment of capital. Overtime these returns became extremely volatile. These include returns for the Bell monopolies that came out of the AT&T divestiture. There was significant increases in variance of the old companies. The volatility in this sector is pronounced. No longer the old regulated monopoly. Lots of risks. Old companies under the pressure of deregulation have changed this game.

1/16/07- Lecture 2: Strategies for Rivalry: Network Sharing Mandates v. Facilities-Based Competition

o Background: Paul Bodrey; Bob Crandallo PBS – Free to Choose in 1980 – discussions of regulatory policy in 10 shows.

- Telecom Monopoly: From Natural to Uno Markets used to be natural monopolies and regulators were okay with this. o 1963-69: microwave long distance – MCI – met regulatory resistance and went to the courts. o 1974-84: US v. AT&T – Antitrust case – complained about the monopoly structure of AT&T.

Allowing things like Hush-a-phone (which was a 1957 case). Illegal to attach systems to the outside of a phone because it was thought to interrupt the network.

1984: Divestiture splitting AT&T up into long distance and 7 local Bell operators that owned the lines.

7 Regional Bell Operating Companies o Line of business restrictionso Equal access for long distance carriers

o 1984-89: Cellular (duopoly) – when licenses were issued, we authorized two licenses in each of 734 markets. Rather than national licenses, they were all local.

- Networks of the ‘90so Local Exchange Carriers (LECs)

Competitive Access Providers – could not provide local to local service, generally. They avoided LECs and attacked the long-distance market. Some became successful in big dense business centers like NY targeting high use of long-distance and bypassing local monopolies.

o Inter-exchange Carriers (IXCs) 3+ nationwide fiber optic networks

o Cable TV systems (CATV) – were originally natural monopolies Direct Broadcast Satellite (DBS) – breaking down natural monopoly

1994 (DrecTV, 1996 EchoStar



o Multiple Wireless (2 cellular, 6 PCS, 1 SMR) – helped break down natural monopolieso Data Networks (Internet) – helped break down natural monopolies

- A New Competitive Paradigmo Competition in long distance and mobile – ppl began talking about the breakdown of natural

monopolieso Multiple Data Networks

1987 Huber Report – the “geodesic network” – wrote a report about this breakdown of natural monopolies. The new technology was putting intelligence farther out to the edges of the network as a result of the reduction in costs of transport (internet).

Computation costs fall relative to transport Routing became sophisticated Applications drive demand

o Cable TV vs. ILECs – just b/c you had a monopoly in Cable TV and monopoly in ILECs didn’t mean you couldn’t have companies creating competition across integrated or converged systems. The idea of “two wires in the home” became popular in the late ‘80s.

1970 cable-telco cross-ownership ban – prohibiting telephone companies from passing video

- 1996 Telecommunications Acto Omnibus provisions – “decency act” thrown out as unconstitutional. o Local Telecoms Goals

Replace regulation with competition (Act promoted competition in local telephone networks)

Create rival networks (competition) Regulation fade away (let markets take over for regulation) But regulate the transition (to boost the competition) – jump start the competition by

helping these new innovative competitive companies to form. Continue Universal Service – universal service is almost exactly opposite of what

you want to happen. These services intentionally drive up prices in some markets so that you can reduce the price below market costs in other markets. Crandall thinks this is bad.

Rural vs. Urban – keep rural markets below market cost by charging more in urban areas. Subsidies are given to high-cost users.

Price averaging Universal service – “we cheat the other guy and pass the savings on to you”

– subsidies get stuck with very profitable rural telephone companies. - Transitioning to Competition

o 1996 Act in Telecommunications Abolish state monopolies – most states (over 40) had effectively state franchised

monopolies. Feds overroad this by establishing a national policy to promote competition. Response of the states was to comply and give out franchises.

Networks obligated to interconnect – established (incumbent) local networks were obliged to connect to new networks. If new networks said they wanted to pass traffic, there has to be a reasonable way to exchange traffic (e.g. between phone networks, for instance). This part of the Act was effective.

AT&T’s control of long distance allowed them to have a premium product. Denial of interconnection was used to block out new networks. This was the rationale for obligations to trade traffic. AT&T denied access by competing phone networks and therefore maintained their monopoly power.

AT&T long lines monopolizationo Mandatory network sharing – seen to be like interconnection, but is different. Here, a

company that does not own a facility has the ability at regulated rates to use the infrastructure of a regulated communications network and then resell the services generated by that infrastructure to retail customers. This was an attempt to boost competition by enabling entrants access to existing networks.



TSR – Total Service Resale – service would be provided by incumbent but new entrant may market to customers. Customers would pay newcomer, newcomer would pay wholesale price to Verizon. Newcomer might add new features and functionality, but essentially the services would also be provided by the incumbent. The wholesale price would be set 18-25% less than retail, providing some margin in there for the newcomer to market the service and make a profit.

UNE – Unbundled Network Elements – making the unbundled components available to competitors to provide competing services. Example is the local loop. This is thought to be difficult to duplicate by new entrants because it is costly to new entrants and must be there prior to selling the service. So it is risky. All these unbundling or resale programs have to include wholesale price controls. If all you say is that the network must be made available, the incumbent could simply say it is available but the price is just way too high. Therefore, you need controlled prices for this to be effective. The Act did not make clear whether this would be permanent.

Note: we have retail price controls today. Voice is retail regulated. You have retail price controls to protect consumers from monopoly pricing. Additionally, new competitors can get wholesale access at reduced rates in order to increase competition.

Local loop, switching, transport, etc. - Policy Induced Competition

o Historic Episodes (Faulhaber 2003) Equipment competition in Bell system. – we grew up in an era where you could buy

and phone and plug it in. In the early days, AT&T owned the phones, you leased the phones, and there was no need for an interface to be able to plug in your own phones. Part of the break-down allowed competition in the phones themselves. This worked because it was technically easy.

Long distance pre- and post-divestiture. – dial in parity. You like dialing parity (1+ and it goes to the long-distance carrier of your choice) – this worked. Simple interfaces worked to promote competition.

o Conclusion: simple interfaces or line of business restrictions needed. The line of business restrictions, meaning that the local Bell companies were prohibited from going into long-distance, wouldn’t try to bias the consumer’s choice of long distance carrier. This made it easier to police b/c the local carriers were limited.

- Stepping Stone Theoryo “Competitors argue that they are making substantial investments in their own facilities and are

using UNEs as a stepping stone to their own facilities.”Gregory Rosston and Roger Noll. Point: these regulations were intended to be a stepping stone, not permanent

regulation. Even competitors using UNEs see them as a stepping stone. - Resale only a transition

o “A possible outcome is that when the dust settles, most local access competition will take the form of resale of the incumbent’s facilities. In this case, consumers are not likely to benefit, and regulation will, if anything, grow as regulators…???”

o Test of the stepping stone theory: look at what happened when we improved the UNE prices so it became a popular form of resale of incumbent infrastructure. UNE-P (Unbundled Development Platforms – got about 50% discounts, making it better than Total Resale Services.

Chart: competitive local exchange carrier (CLEC). CLEC-owned is like a… UNE-P got big in around 2000. As UNE-P gets big and forces TSR to flatten.

Comment: if the UNE-P was just a stepping stone, you would see these new CLECs you should see more CLEC-owned, but we don’t see this trend. As resale lines takeover, CLEC-owned lines remain flat. If you take cable out of CLEC-owned, there is actually a decline for a three year period. You are actually dumping CLEC-owned lines. You aren’t getting the stepping-stone to CLEC-owned lines.

Big issue: cable. Cable was already in place. The incremental cost of putting phone service would not be high. Why aren’t they doing this? The



incentives aren’t right. Their costs are enormous to get two-way broadband. It is a result of regulation: retail price constraint keeping prices down. Moreover, you are going head-to-head with a telephone company that can simply lower the UNE-P rates. If you are a cable operator, you have to be very conservative about putting $1-2B investment when that can be flattened by regulators.

In 2004, federal courts found a violation in the 1996 Telecommunications Act. Judge Williams says they are after the head-to-head networks. If you have too much resale, you will undercut investment in head-to-head networks. Therefore, it was thrown out in 1999, 2002, and was thrown out in 2004. The regime collapsed between that opinion in March and June or July. The UNE-P rules then went flat, and are headed south. They are going back to the incumbents or to CLEC-owned. Now, there is a slight kick-up from the cable lines. Cable telephone coverage at the end of 2003, prior to collapse of UNE-P rules, only covered 15% of US households. The number in Q3 2006 says that 69% are now served. This happened between 2004 and Q3 2006. It has taken off like a rocket.

- Crandall takes over lecture…- “Stepping Stones” or “Stumbling Blocks” – Mandatory Network Sharing in Telecom

o Background: UNE-P – deals struck for; long distance companies had incentive to get into local service to remain competitive. There is no evidence on the record that by using a UNE-P, despite 50% discounts, that they made any money on local services. This is because it is very costly to market local telephone service. Getting someone to change their local service to save $3-4/month is difficult. Therefore, the marketing costs were huge. The long distance business was going down rapidly because of the increased competition in the market, as well as the competition from wireless carriers. Wireless began to take over the business. The long-distance carriers avoided going out of business through mergers. Verizon bought MCI. The idea of an independent long-distance business is dead. The 1984-divestiture is really just a historical artifact at this point.

o What led to unbundling in the first place? In many countries, the bottleneck was never a “monopoly” bottleneck, just an

expensive one. Unbundling and network sharing are regulatory interventions of last resort where

there is not a second, third or fourth network providing access to the same households or establishments.

These network sharing arrangements were initially thought to be temporary – i.e. “stepping stones” which could be abandoned once entrants built their own facilities.

Today, network sharing appears to be a permanent fixture in the US, Japan and Australia.

There are always new networks being created: today, it is WIFI, Satellite – KA-band satellites that can provide internet to everyone in America not on the Northside of a mountain, etc.

o Rationale for unbundling was to provide a stepping stone. However, cable companies never used these stepping stones. They are getting in through VOIP, etc. The US, because of a court case, and Canada because of a less aggressive policy, are not going after network sharing as aggressively as in other countries.

Other countries: - Has the Policy worked? Are the net benefits positive?

o Presumably, the objective was to accelerate competition which, in turn, should reduce prices and/or provide innovative new services.

o If one believes the “stepping stone” hypothesis, the policy should also induce investment by entrants in new facilities as they step across the stones, climb the “ladder of investment”, or whatever…



Reality: everyone already had local telephone service. All these regulations were basically just about reallocating money from ILECs to consumers. Maybe a valuable goal, but regulation is an expensive way to do this.

o But any regulatory intervention of this magnitude has offsetting costs: it reduces the incentives of the regulated (ILEC) firm to invest, innovate, and deploy its own new services.

o So what is the evidence on these matters? - Did the Policy Help Create Competition in narrowband services?

o Unbundling only tried in the US and Canada for narrowband, voice service; clearly a failure in both countries, particularly in the US.

o US CLECs reported capital expenditures of more than $60B. No net gain in the residential market. There may have been some benefit in the business market. (perhaps $.5B/yr, if you assume there was no technological progress). There is no net gain in residential, and very little gain in business. A terrible failure compared to deregulation in other markets like airlines, energy, trucking, etc. This comes about because lower prices lead to more choices and more competition.

o Here, none of the CLECs really made it, they invested $60B and lost basically all of it. There is very little evidence that this competitive entrant experiment for local telephony provided any real value.

o Court ruling: Regulators cannot mandate that they have to unbundle everything in every market. It may be required in rural parts, but not in major cities. This basically destroyed MCI.

o Cable is now entering the market with VOIP. - Current Test of the Stepping Stone Hypothesis is Broadband Services.

o Broadband is where the game is. In June 2006. o Chart: DSL (from ILEC), DSL-network sharing (by entrants using incumbents network

facilities), Cable and other broadband options. Breakdown within countries. Trends: where cable got a good grip, there was more DSL and more cable. Good

competition between the two. In others, there was less competition in terms of broadband penetration.

Question: does DSL-network sharing increase total broadband penetration or does it simply substitute for other DSL and Cable use.

- Is Local Loop Unbundling Necessary to Promote Broadband Competition? o Regulatory logic – Competition in “DSL Market” is required to limit incumbent telephone

company monopoly. o But there is no “DSL Market” – only a broadband market in fact in U.S. DSL is far behind

cable TV in broadband race. o Local loop unbundling (LLU) is generally not attractive to DSL competitors. o Entrants want “line sharing” i.e. a lease of only upper frequencies of the telephone line. o US abandoned line sharing and unbundling for DSL after 7 uears of regulatory battles

between telephone companies and new entrants. o ….

- Substantial capital investment required to upgrade network infrastructure to deliver high-speed services

o Japan and Korea have 20-30 Mbs broadband because of substantial population density and short substriber lines. (shorter loop links).

o US and EU services are often only in the 2-6Mbs range. o Iliad (Free reports speeds of up to 20Mbs. o US telephone companies are forced to spend billions of dollars extending fiber optics into

their networks, but likely will not spend it if required to unbundled this new capacity for entrants.

Less incentive to do this if there are wholesale regulated prices for resale and mandatory line sharing.

- Investment incentives and network unbundlingo Sunk costs and investment



o Investment in new broadband capcity is “sunk” i.e. once deployed, it cannot be removed and used elsewhere.

o This investment is also risky because new technology may render it obsolete in a few years. o “Unbundling” or “line sharing” at cost-based regulated rates limits incumbents returns if

technology proves successful. o But if DSL technology becomes obsolete quickly, incumbents cannot lease it at any price. o Therefore, LLU …

- But Few of the Companies in North America, Japan or Europe are Climbing the Ladder of Investmento Exceptions are that once you allow unbundling, they do not create their own facilities. The

exceptions are Illiad (France), Wind (Italy), Tiscali (Italy), Talk America (Michighan). o Competitors in Japan …o …o …

- The Cost of LLU and Line Sharing: Less Network Investment? o Chart

- EU Has Had Much Less Network Investment than USo EU has pushed unbundling and network sharing which is preventing investment in new

networks. Fixed line investments seem to be decreasing between 2001 to 2004. US companies have been investing far more in total, even though the EU is 375M people and we are around 290M at the end of 2004.

o US spends substantially more than the EU on network lines. - What are the EU Regulators’ Views of Telecom CapEx?

o The European Commission’s 2006 implementation report asserted “- Unfortunately, there is no evidence that network sharing increases broadband penetration

o …find that platform competition is much more important in driving broadband. o No studies have found a strong, long-lasting effect of network sharing on broadband

subscriptsion. Europeans have no evidence that their policy is working.

o Exceptions may be France and Japan, but DSL seems to be slowing dramatically in Japan. The next battle is over sharing NTT’s fiber to the home (FTTH) network.

- Conclusionso Seems to be deleterious to network investment. o Does not seem to increase broadband penetration. o The ladder of investment is missing the top rung; Entrants are not building their own

facilities. o No evidence that network sharing increases broadband penetration or otherwise creates

benefits for consumers. o There is at least preliminary evidence that network sharing has reduced network investment

by the incumbents. o Despite these results, there is good news: these policies create rents for lawyers and

economists everywhere!

1/23/07- Telecom Network Sharing Mandates/Cable TV Regulation- Wrap up: Network Sharing Mandates

o Regression analysis: testing the stepping stone Hypo: resale lines in period t create CLEC-owned lines in period t+1, t+2…

o Investment flows: suppose you allow new entrants to rent facilities from incumbents, and that forces prices down. Isn’t competition always good? If all we’ve done is create retail competition right now, you can’t just add up the cost savings and say that is the benefit because the incumbent may not be able to invest as heavily for the future benefit of the customer.

Isolating effect of macro variableso Policy positions of vested interests



Interested parties favor rules that increase equity value. - Efficiency is a marketing efficiency and start up cost. The philosophy is to try to encourage the entry

by the resale program and thereby encourage new entrants to slowly develop and create their own lines.

- Interested party commentaryo CLECs want low UNE rateso ILECS oppose low UNE rates

- Consensus on Wall Streeto UNE-P isn’t competition as the market defines…

- “Bright spot” in 2003-Io “practically every telco reported capex well below their expectations.” – they were happy on

Wall Street that phone companies were taking money out of the companies. As advisors, they were saying don’t buy shares of these companies until they stop investing.

- Consensus in Manufacturingo Aggressive unbundling and pricing rules can create perverse economic incentives for

competitive telecommunications carriers to rely on the incumbents’…- Consensus in Silicon Valley

o Dr. Grove stated that Intel’s goal is to make true broadband…He was against unbundling. - Consensus of Broadband Providers- Cable TV as MVPD

o Market powero Prices (video) above average costs

Prices, above marginal costs necessary to recoup sunk costso Evidence

Prices in duopoly markets 15-20% less Q ratio = mkt value/replacement cost of capital

o Will price regulation lower prices? - Regulation and market power

o Burden of proof key Marketplace? Regulation?

o Market failure necessary to justify regulation But insufficient Ergo, regulation properly bears the burden.

o NOTE: the regulation may do more harm than good. It is not clear that the regulatory framework will reduce prices for consumers. Therefore, the burden of proof needs to be on the regulators to justify what they want to do. Look at how competitive the market is as a predicate – if you do have a monopoly, that is the beginning of the inquiry, not the end. The End is justifying the benefits and costs of regulation.

- The Cycle of Regime Switcheso Retail rate control regime switches

Jan 1, 1987: Decontrol (federal pre-emption) May 1, 1993: Re-regulation (1992 Cable Act) Nov 10, 1994: Decontrol (‘going forward’) March 31, 1999: Decontrol (1996 Telecommunications Act)

- Two Problems, One Resulto Difficult to lower nominal rates

Market prices indicate willingness-to-payo Impossible to lower nominal rates and yet prevent quality adjustments by operators

Δ tiering (“mini-pay” tiers) Δ programming costs (fireplace channels) Δ marketing efforts (basic vs. premium) – change marketing from the regulated

(basic) to the unregulated (premium). Δ investment flows



o The ‘problem of free speech’? Government cannot assert authority over the content. They can’t say the fireplace

channels are out. That is a violation of the free speech of the cable operators. o Net result: output not increased via controls. Rates were not effectively controlled

- Politics of Regulationo Rate regulation supported by broadcasters

NAB funded 1992 ad campaign by CFAo NAB’s (national association of broadcasters) proposed for $4.52 price capo Telcos favored cable rate regulation. o Programmers lobbied against controls (and succeeded in ‘going forward’)

1/30/07- The Wonderful World of Wireless

o 1895 - Guglioni Marconi – a Brit/Italian, inventor of wireless Wireless communications were introduced. Spectrum scarcity. Wireless

communications introduced scarcity. Prior to radio, there were no applications that made the frequency space important. When the applications began appearing (e.g. radios) it became valuable to engage in wireless/radio communications.

Advances in the technology categorically raised demand for spectrum access. On the other hand, there was simultaneously some altering of valuation across bands. If you have a really efficient radio technology, it might reduce demand to use certain radio bands because the new system might be very productive at using a particular bandwidth and make it less interesting for consumer’s to use other bandwidth. These valuations then fluctuate in complex ways in response to technology.

Initially, there was an open access property regime. A classic example of a new discovery of a valuable natural resource.

o Valuable applications emerge thanks. Titanic. David Sarnoff, founder of RCA, heard the SOS for the titanic and relayed it to some folks in NY. One of the first applications of radio was ship-to-shore communications, particularly in emergencies. With valuable applications being demonstrated, the ability for conflict emerges.

1912 Radio act – licensing regime for point-to-point (e.g. ship-to-shore, etc.) Net benefits of coordination (property rights) increase.

o Commercial radio broadcasting 1st killer app. If you have point-to-point radio, when you have a beam going from one point to another, you aren’t trying to spread the beam out (this wastes energy). However, in 1920 the first commercial radio station of America emerges. KDKA in Pittsburgh. They broadcast the election returns in the 1920 election.

Strategy: who builds the first radio station? Why is this an interesting question? You can’t make money on a radio station until people have radios. You can’t make money on radios until you have radio stations. These are classic tied goods.

How did the regulation go? The regulatory regime was surprised by the new application of commercial broadcasting. KDKA was launched with a point-to-point license. There wasn’t anything called a broadcasting license. KDKA started broadcasting things like symphonies, election results, etc. KDKA was the first company that decided to have an ongoing broadcasting business.

1922 – FCC decided there should be a broadcasting license. Under the 1912 statute, they had no way to refuse a license excepts on grounds of interference. By 1922, there were 500+ radio broadcasters. They reacted at the Department of Commerce by passing out licenses. They are trying to coordinate the market, keeping people in cities off each other’s frequencies. They experimented with timesharing, etc. 1924 was radio Christmas. By 1923, the radio dial expanded its frequencies. However, the D of Commerce was doing this with one hand tied behind their backs because it was hard to refuse someone a new license.

1927 Radio Act – signed by President Hoover. The major commercial broadcasters actually did not want a system of property rights. D of Commerce already adopted



common law rules. There was a “priority in use” system. If you got in and started broadcasting, you were protected from newcomers. This has many names. Right of user. Adverse possession. Squatters poverty. Right of first appropriation. This was a system being enforced by the DoC but the incumbent radio stations didn’t like it (ironically). In 1925, they created the National Association of Broadcasters (NAB) and argued in favor of a property right system. Hoover wanted legislation almost from the beginning. Hoover stops enforcing the law. On 7/9/1926, he issues a statement from the DoC saying they aren’t going to enforce the “priority in use” rules. 200 new stations came in. And there was a period from 7/9/1926 and 2/1927, in which it was called “The Period of the Breakdown of the Law.” Every radio station could do what they wanted.

- Spectrum Allocation Policyo “public interest” allocation.

Overturned “priority in use” DoC rules Oak Leaves during “Period of the Breakdown of the Law.” Judge looks at

the case. Cook County judge says we understand the concept of property rights in the “ether”. He says there are lots of things that are intangible that have property rights. (e.g. water law). WGN built up a good audience, investing in broadcasting equipment, creating a product, and then someone moves next to them and hurts their ability to conduct that business. Judge provided WGN that they have a property right and that Oak Leaves has to go 50kilocycles on the A.M. dial in either direction within 100 miles of downtown Chicago. Legislature thought it might be a problem that judges would start creating property rights in the courts. Within three weeks, there was a resolution passed by both houses and signed by Coolidge. There is a very political reaction to this. The folks that want a centrally administered system, including the NAB, move quickly to get a law passed about 80 years. (2/3/1927).

NAB proposed “public interest” – Why would the incumbent broadcasters want a public interest property right

system rather than. No regulatory system would appropriate the investment the incumbents already had made. Therefore, they thought they could assert the defacto property rights of the system but it would be more difficult for competitors to enter the market if there were public allocation interests (e.g. property rights). Therefore, these regulatory rules that were favored by the incumbents would prove fatal for the small competitors. Broadcasting now has a political edge to it. The regulators and policy makers saw that this was going to be a very important medium of opinion. The policy makers that would have to run for public office saw radio and television stations as creating that output that was a major input into a “political production function.” The regulators and policymakers want to be in some kind of quid-pro-quo relationship with the broadcasting groups.

Equal time rule – in political debates, you had to give equal time to challenger and incumbent. Incumbents are always in favor of things like the “equal time rule” because in reality they get more as the incumbents and they go in ahead anyway.

o Mobile telephony eclipsed – we don’t see the same political interest in having tight restrictions and micromanagement in the regulatory environment as we had in broadcasting. What happens is interesting. Public interest of radio spectrum does not explicitly (or implicitly) allow people to own radio space. You are given a license. You are given the right to operate a transmitter at a particular location using a particular standard (NTSC – national television standards committee). You can’t take your spectrum, go dark, and then change from television to mobile. You have control over the TV station and the type of programming, subject to certain restraints. That is a very restrictive policy that has been



criticized as a command-and-control regime. This political interest in broadcasting is what really brought TV and radio together with regulators and have very little competition in the markets. The fact that we are watching the same channels that were planned for production in 1941 is a joke!! Isn’t it? You get this transition to the cellular world where the political influences are not as well justified.

2001: $17.7 billion revenue terrestrial radio 2001: $54.4 billion broadcast TV 2001: $65.3 billion mobile telephony

- Origins of Cellularo There is a new world created by the economic reality of wireless telephony that has little to do

with the old political bargain (e.g. policy makers saying they will pretend to regulate the broadcasters in exchange for getting very limited ability to compete with each other but the market will be nice and protected). This seems to work. Both Democrats and Republicans like to regulate content. Republicans don’t like sex, Democrats don’t like violence.

o Hazlett has a 1990 Article about all of this…It is one of his favorite topics. o Radio phones date back to the 1930s

One phone call, one frequency, one market – the radio common carriers (RCC) market. You could only have one user of a market at a time.

o Cellularization: less is more. Divide market into cells. Rather than having one frequency serving all of NY, you

chop that frequency up into 50 cells and put a base station into each cell site and use low power for the base transmitters. Low power because you are only trying to send a signal locally rather than all the way around the city.

Lower power used by radios to transmit. Reuse cells, multiple bandwidth. This is geographic reuse. You can multiply your

frequencies. Geographic reuse is far and away the leading method for increasing spectrum capacity or usage.

Hand-off calls from one cell site to another for mobility. High speed hand-off happens anytime you are talking while driving in a car, for instance.

- Historyo Bell labs conceived cellular in 1946. o FCC did not start proceeding until 1968.

Upper UHF bands proposed. Channels 70-83 (14x6MHz = 84 MHz) of bandwith is taken from the TV

bands for this. TV broadcasters resisted this, obviously. AT&T did not push hard for this. Why would the monopolist not push hard

for this new service? They thought it was iffy whether it would be a killer application and AT&T owned the market. They saw it to be cannibalization of their own market.

FCC proposed a cellular monopoly, and assumed it had to be a national monopoly tied to AT&T, the wired monopoly. This was the official view government view in 1968. We have clearly proved this false.

- Duopoly was Radical Reformo Rulemakings up through 1982 introduced competition as a wild idea. o FCC settled on 2 licenses in each market

One to local ILEC, another to an entranto 734 local markets

Licenses issued by lottery. In the lotteries, the big markets went from 1984-1986 (306). The rural areas went from 1988-1989. The lotteries required that you be a “phone company” in order to get a license. (428). You got one draw in 428 markets. You would pay $300k. usually, you went in with 40-50 other companies so each company paid a few grand.

o 25 MHz per each of two licenses. o Analog standard mandated: AMPS. Therefore, you did not have a choice of what to provide.

In 1988, FCC repealed this and allowed free choice in the type of technology used.



- Surprise. It Workso Technology feasible. o Investors support the infrastructure. Cingular has 45,000 base stations in their national

network. You have to have investment in customer service and coordinate with handheld makers, etc.

License values increase monotonically in 1980s. By 1990, total value of licenses was ~ $100 billion.

o Demand for wireless more robust than initially thought. - State Rate Regulation

o States allowed to regulate rates Retail rate regulation Wholesale rate regulation – government says that a network (say Cellular 1) has to

offer its network infrastructure to competitors. This is very much like unbundling or resale in the fixed business.

Enabling resellers Precursors to MVNOs (mobile virtual network operator) that sell you a

phone and prepaid minutes. They buy billions of minutes wholesale from Verizon, Sprint, Cingular, etc.

- Pre-empting state regulationo 1993: Congress pre-empts state rate regulation pending 1-year phase outo States may request continuing authorityo 7 states filed petitions with the FCC

Burden of proof on markets, not regulators (states argue). Burden of proof should be on rate regulation, not on the markets. Typically, federal regulators position is that if someone is going to mess up the market it is going to be them. Economic interest of regulators – they want to be in control. Therefore, federal regulators will overrule local legislation to stay in control.

o Federal pre-emption of state rate regulation effective August 10, 1994- 1994: Deregulate Cellular?

o Natural experiment – feds deregulate. What should happen? Regulated cities (boston, LA, NY, SF), Unregulated (Chicago, Dallas, Washington

DC, Philadelphia) You would expect rates to fly up and growth be restricted in the regulated subset.

Unregulated penetration is higher to begin with. There is no evidence of any positive effect of rate regulation.

Rate regulation did not make the markets better. Finally, competition comes to the rescue.

Prices under the old duopoly and regulation in many states was around $0.50 and even went up. Prices drop substantially when licenses are auctioned, etc. There is an 80% drop in the nominal price (not adjusted for inflation). Prices now are under $0.07/minute. Minutes of use has skyrocketed.

1998 is the year that usage really starts to take off.- Competition and Consolidation

o Demand for mobility + economies of scale efficiency of national networkso Scale achieved by

Merger Roaming agreements Standardization of technology (equipment markets)

- Wireless Market Concentration and Avg Rev per Minute (1995-2004). - Organizational Complexity of Wireless Networks (WWANs)

o De facto private ownership, management of frequency space. As an economic matter, regulation in the US where we don’t regulate the technology. The US allows the operator to have exclusive authority over the space with very liberal rules over the models they use (subscription models, for instance). You can choose your technology. We’ve deregulated all that stuff. You can provide data service, voice service, video, etc. Prices and quality.



o “Exclusive use” intensely shared. The exclusive ownership of the airways. This is a stupid term. Exclusive ownership makes more sense. Mr. Verizon is not the only one using the spectrum.

- Spectrum Sharingo Multiple layered netowkrs (analog, digital, voice, data, EV-DO, EV-DO Rev. A). o Blackberry has a network of servers but, by K, their use is given to them by Spring, or

Verison, etc. - Blackberry, On-Star, CDPD, etc. o Wholesale MVNO markets (e.g. TracFone)o Roaming Agreements – precreated Ks by Sprint or Cingular giving you 150 more networks

internationally, that are wrapped up in subscriber agreements. o Equipment makers (e.g. Motorola) get to share the spectrum. o Application vendors (e.g. DoCoMo, pix) – wireless web services have been set up and

applications are available through a mobile phone. They tell Yahoo and the application providers exactly what to do to optimize their presentations, graphics, etc.

o Subscribers – spectrum sharing among subscribers. All that optimization is totally seamless to you. The network coordinates everything.

2/6/07- Burden of Proof

o Is the market competitive? vs. Is Regulation Effective? o Market not “perfectly competitive” implies…what? There are no competitive markets in

which products are perfect substitutes for each other. Prices should not equal marginal cost. This is what is taught in economics courses.

If firms in perfectly competitive markets charge prices above marginal cost, they will be pushed out of the market. This isn’t the case in real markets. All markets have sunk investment in order to provide efficient services. If you only charge marginal cost, there is no compensation for the infrastructure. You never get to recoup your investment. In a world that is forward-looking in which people look forward before investments, if they don’t expect to recoup their investments, you will not have efficient levels of investment. Services consumers would pay for would not be provided because they would require common pooling of investment that could not be recouped by a firm pressured by regulations to charge only marginal cost.

Therefore, prices should not be marginal cost. Prices efficiently capture fixed costs

Inelastic customers pay more. These customers will not restrict their purchases just because prices go up. Therefore, prices should be significantly different for different types of customers.

Customers supporting the network…- “Workably Competitive” Theorem

o “To determine whether any industry is workably competitive, therefore, simply have a good graduate student write his dissertation on the industry and render a verdict. It is critical to this test, of course, that no second graduate student be allowed to study the industry.” – George Stigler.

o Even if it is not perfectly competitive, will regulation help? In many cases, the answer is no because of the costs involved. This doesn’t have anything directly to do with whether the market is perfectly competitive or not. Therefore, the key test for the effectiveness of regulation is whether that regulation increases consumer welfare.

- Apples to appleso Ideal regulation vs. imperfect markets is not the appropriate assessment. Regulation cannot

costlessly fix whatever is happening in the market. Moreover, the opposite is also true. Flawed regulation vs. perfect markets.

o Assymetric analysis irrelevant answerso Appropriate Question: Actual test: does actual regulation improve performance of actual

markets? Performance in consumer welfare terms



Accounting for current and future costs, benefits Current retail prices, investment in future services are both important.

Incentives for future innovation and compensation in future periods. It is possible that regulation can reduce prices in one period while really screwing up the market in the long-run. A one-period assessment is not the appropriate one.

- Forbearanceo TELUS (local operator with long distance and other operations – more entrepreneurial) – they

came up with this concept of “forbearance”. They went to the regulators with a proposal drawn up by company lawyers.

o Forbearance: we deregulated long distance in the late 1990s. This was when AT&T was the major carrier. There were regulatory rules preventing AT&T from lowering its prices to allow MCI and Sprint to have reasonable ability to make money. We moved to market pricing in the 90s.

- Thinking the Unthinkableo Deregulating local retail voice telecoms

Long distance deregulated (late 1990s) Wholesale voice deregulated (2004) – unbundling, etc. It is still possible to rent

local loops (last mile wire) going into customers premises. However, no one is doing this because the rates are too high as a reseller.

Wholesale broadband deregulated (2005) Retail cellular deregulated (1994) VoIP rates unregulated (e911 imposed) – these rates are not regulated by any state or

federal agency in the US. There is some regulation (e911), which is quite a burden. o Cold Turkey for POTS (plain old telephone service)? Can we use the same approach to retail

rates of regular telephone service. - Rival Network Platforms

o Many networks are bringing information into the home. Cable, telephone, satellite, wireless networks, electric wires

o These rival platforms are coming. o Forbearance is arguing that the world has changed and there is “intermodal competition”.

The argument is to get the regulators out of the picture and let the markets work. (cable operators offer basic telephone services to 70% of households. 95% of cable offer broadband in addition to video)

- Wirelesso Excellent study on its own

Organization of markets (network operators with licenses to use spectrum and build complementary infrastructure). There are a lot of business Ks. They organize markets in a very complex way.

Challenges for regulations – o Directly relevant to emergence of “facilities-based competition” in telecoms. Wireless

represents a number of networks that are already in place. Most advanced countries have at least three wireless networks.

1996 fixed line services were dominant. This was the framework of the 1996 Telecommunications Act

2007: mobile services are highly competitive with the fixed line services. - Cellular Phone Markets

o Cellular was initially thought to be a business service (as recently as 1995 and 1996). o Difficult to regulate

Even with duopoly prices > cost. We define the market as a duopoly. We observe the prices of the duopolist were incredibly high. The value of licenses to be a cellular operator in 1990 was about $100 billion. That was the capitalized value of the profits being earned.



State regulation experience – unable to lower prices. Prices did not go up in places where it was regulated. Control group were the states that did not have any regulation before or after. We’ve seen how difficult it is to regulate.

o Markets organize complex transactions Billions of spectrum access ‘purchases’ Consumers buy minutes of access to networks and spectrum. Every time you may a

cell phone call, you are buying airspace controlled by the network and the complementary inputs (e.g. the fixed infrastructure that works with your particular handset). These phones, services and networks are bundled. The costs of unbundling are considerable.

Phones, networks and services are bundled – it is never true that firms try to extend themselves to provide an entire vertical chain to the customer. They always have to purchase some external inputs. The firm has to stop somewhere.

Competition governs quality, coverage, customer service, and “interference”o Liberal rules in the USA – prepaid, postpaid models; bundling;

But, there is constrained spectrum allocations. - Simple take on Complexity

o 2004-Q1: AT&T Wireless bought by Cingular and &&& bought by Sprint. o T-Mobile uses GSM. o Nextel has “iMode” which is a Motorola invention.

- Market Feedbacko “churn” stats drive share.

Reflect everything that matters to customers. It is very expensive to get a new customer. Price, coverage, call quality, applications and functionality, handset selection,

customer service and billing. o Simple but effective policing mechanism

2004: US wireless carriers spent $28 billion on network infrastructure and upgrades 2006: ~200,000 cell sites. Cingular has 45,000 cell sites. These are about $500,000

and up. o Financial analysts – will compare growth in revenue, how much they are spending on

equipment, operating costs. Cingular’s churn is 1.8%/month compared to Verizon at 1.14%. This equates to 1.3M customers they have to recruit each quarter just to maintain the status quo. This means they are extremely sensitive to dropped calls and bad customer service. These are hard problems. Verizon is even disclosing more numbers on churn, making their operations even more open. Full disclosure. This is pushing Cingular to do the same.

o This market is a perfect example of where competition is beating regulation. Whenever competition is beating regulation

- Cingular v. Verizon Wireless: Battle of the Titans- Consumer info – helpful but…

o Which network is best for you? o

- Thinking the Unthinkableo Forbearance: legalese for agency deregulation. The biggest reason a company like TELUS is

going to regulators and arguing for deregulation in retail regulation and wholesale regulation. The agency can do this simply by forbearing. The regulator just says they looked at it and they decide not to regulate. You don’t need a statute. They can forbear when they think it is in the public’s interest.

o If retail price controls are doing their job, you don’t need wholesale price controls. It was reticence by regulators. They didn’t want prices to fly up, and they wanted to also try to keep prices down from wholesale regulation. The fact is that wholesale is basically ineffective in Canada and has come off in the US.

TELUS: told CRTC that they should forbear in telecoms and proposed a bright-line two-part test: 1) competitive network facilities are in place (that means there is actual



infrastructure in the ground for a competitor to provide voice service head to head with the ILEC, and 2) that new competitive network has > 5% of market share.

Why only 5% market share? 1) has someone invested significant capital and sunk it? If someone’s sunk capital to go head-to-head, it is not going anywhere and can’t be redirected to other use. It is in the owner’s incentive to make a go of it. 5% is just a breathing test to make sure it is a real operator offering a real service going head-to-head), 2) what are the duplicative or rivalrous facilities we’re talking about? What are the competitive options? They are thinking about a) cable television, and b) cellular options. Once those networks are in place, you don’t have just one competing network, you have multiple competing networks. The Canadian government actually accepted the argument with a 15% market share test.

The baseline is that there is not a monopoly, you have competition. How did the regulators respond to this?

- Key Argumentso Regulated network sharing did not work

Administratively difficult to arrange “forced marriages” – you’ve got an incumbent network, you are making government rules about how another company can use that private network to compete at retail. This is a forced marriage of market competitors because they have to share the same facilities. There were rules and disputes over what access to the bathroom in the central office you had to give the technicians of the competitors working on the network. You want these companies to compete. Why wouldn’t they prevent their competitors from using their bathrooms? Voluntary marriages are difficult, not just forced ones.

Attracts rent-seeking – a lot of money put into lobbying regulators and legislatures, and this is incredibly costly. The term is “wasteful rent-seeking”. Lots of expenses in rent-seeking

Undermines incentives to build and/or innovate – this was in fact the case. This disincentive went over to the entrants. We had less facilities based entry. When we cut out wholesale regulation we got more CLEC-owned lines.

o Price controls distort markets Undermine efficient investment (if too low). Thwart entry (if too high) – entry won’t take place if you set price controls too high.

You defeat the whole purpose. Can perversely achieve both (given uncertainty about future price levels) –

we might have high wholesale prices that have a level of uncertainty about them because they can change at the whim or regulators. As an investor, you don’t want to invest in either the incumbents or the entrants systems. There are all sorts of disincentive effects for investments that can be problematic.

Limit pro-consumer competition (price floors) – - Wrath of Kahn

o Kahn is a regulatory economist. He wrote a book in 1970. Kahn worked for the Carter administration. He gave a talk as the Chairman of the Council of Economic Advisors. He was on the state public utilities commission.

o Why is Kahn so mad? UNE-P priced at TELRIC. Wholesale rate regulation is certainly not over. The

OECD believes this is the way to go. What happened in the US, is that the price MCI or old AT&T could rent the Verizon rates were set at “Total Element Long-Run Incremental Cost (TELRIC).

Telco entrant leases ILEC network at the incremental cost of an efficient new network – the FCC allowed the entrants to lease or rent the facilities at rates that were determined hypothetically by what the incremental cost of the most efficient network would be, right now.



TELRIC-BS (blank slate) – that allows the regulators to set a rate much lower than even the incremental cost of the existing network. The regulators were basically saying, we don’t care what the actual prices actually are. We’ll set the price at what is efficient. The fact that they had higher costs because sunk costs are sunk. Kahn is mad about this. The regulators seem to think that by analogy to perfect competition they were setting efficient prices because sunk costs are bygones. However, it is a bad application of a very limited theory.

o Theory: sunk costs are bygones. In a forward-looking sense, if it is known to investors today that their investments will not be recouped, you’ll get less investment today. So you’ll get lousier services, and fewer of them, at higher prices.

- Three Problemso Dynamic technical change

Costs decline over time – why put this money in the ground if every year in the future you new there would be better technology and lower costs and you would be forced to lower your costs to meet these. In real markets, people invest when they get to make a hefty margin in the first few years. We took that away with both retail and wholesale price controls. They couldn’t brace themselves over the longterm with their “old” technology.

o Incremental costs may not cover fixed costs when AC > MC (often in telecoms)o Option value of waiting to sink capital may be substantial (reduces standing risk) – option

value is when you delay decisions. It is important in the marketplace. We pay money for a longer period of having the option. If the regulator steps into this world and tells all Telecom investors, put the investment in the ground today but 2-3 years from now we will price all your capital at then prevailing market prices for then efficient systems, that significantly increases the option value of waiting. All this uncertainty about the new technology makes it even more costly to shoulder and make it more likely you will want to wait.

Undermining incentives to invest. Consumers are not getting the investment and the services that result because the risks and the appropriation (inability to recoup sunk costs) is deterring investment from investors that can put their money elsewhere. For the entrants, they got basically a free-ride. They were discouraged from creating their own infrastructures because they basically get this option for free. If they are the entrant leasing the facilities, they get to come in, leave if they don’t like it, and that is an amazing advantage. Go look at office space in the DC market versus a two year contract basis, and the differences are astounding. There is at least a 4-1 differential on office space. Why is that? Option value. On a month-to-month basis you have the option to leave.

All this deterred incumbents and entrants from building and improving facilities. Kahn thinks this is a standard problem and we should look at what happens when you have a high fixed cost, low marginal cost market. This is basically all the new markets. Google: high investment in fiber networks. Low cost in services. Intellectual property: high fixed creation costs; very low marginal cost for creating additional copies.

- Efficient Prices in Network Inidustrieso Key: FC High, MC Lowo Must, looking forward, reflect full compensation for investors (covering sunk costs)o Must compensate for option valueso Otherwise, investment will dry up in the sector (both incumbents and entrants)

- Pricing with Common Costso Irony, P>MC inefficient…but necessary to recover common costs. On the one hand, we

understand the theory in the text books that this is inefficient because the opportunity cost of using more of the network is zero. However, if we want investment, how do we encourage?

o Solutions



Multi-part tariffs: traditionally in telecoms, people pay a monthly fee (fixed charge). This is like paying to be a member of the club. Additionally, you pay another fee for the amount you use. This is usually around marginal cost.

Ramsey prices The price markup over marginal cost for good n is inversely proportional to

the elasticity of demand. You are marking up costs for inelastic customers. You’ll have a low price if highly elastic, and a high price if inelastic. Charge the business flyer more. Charge the senior citizen bargain shopper less.

Recovers common costs, etc. - Textbook Meets the Real World

o Neither pricing strategy in “perfectly competitive” marketplaceo Deviation is a product of

Efficient network services?- Efficient Prices do not equal Marginal Cost

o Must be expected to recoup sunk outlayso Must be higher for inelastic demanders

Discrimination against business air passengers Discrimination against wealthy med. Patients Diescrmination against early EV DO adoptors

o Comeptition forces such pricingo Regulatory dilemma…

- Risks of Forbearanceo Political Risk: Retail prices could immediately rise o Risk 2: retail prices could fall.

- Price Deregulation Fly-up? o Is it likely we would have a price fly-up after deregulation? Crandall would say “no” because

we have competitive networks. Multiple mobile networks – imperfect substitutes (mobility v. QoS). Residential vs. business – different demands. Business wants the QoS.

o VOIP Close substitutes (and not so close) – skype is cheap Cable modem as independent VoIP network. Fiber overbuilds in business markets

o Local market coverage? o US retail phone rates declining with deregulation

- Real Residential Fixed Line Phone Bills: 1995-2005. - End of UNE-P

o No price fly up- Inter-modal Competition

o Historically important – regulators did not spend enough time thinking about the possibilities of technological change.

MCI entry against AT&T long lines Fiber optics in metro bus areas Cellular against “fixed” line Cable TV system broadband service

o Highly disruptive Coordination among rival companies

- Predationo Standard antitrust violation if anti-consumer. Incumbent firms will use predatory terms to

repel competitive entrants. Once the competitive entrants have already built their network, lowering prices to drive them away no longer works. This is a preemptive strategy.

o Not profitable if rival networks are in place Important not to protect rivals from pro-consumer competition



o Empirical observation Cable TV pricing in USA

MISSED CLASS - LECTURE 6

Lecture 7: Can the Internet be regulated? - Deregulation-enabled?

o What did gov’t want to deregulate? Information services or enhanced or advanced services. This is the stuff you want to deregulate. The stuff you want to keep regulations on are basic, telecommunications, or common carrier services. This is the old stuff.

o When it comes to the old stuff, people tend to be more willing to regulate because the cost of investments (laying wires, etc.) have already been incurred for some companies. However, for new technologies and services, we want to encourage innovation through deregulation. This is basically the regulatory side to the internet, etc.

o Jesse Oxman: “The success of the Internet ahs not been an accidental development. Market forces have driven the Internet’s growth and the FCC has had an important role to play in creating a deregulatory environment in which the Internet could flourish.” We could have crushed it with regulation and we didn’t – give us credit.

- Lesson’s Learned (Oxman)o Refrain from “legacy” regulation – don’t regulate the new stuff like you regulate the old stuff. o Deregulate the old but after competition emerges. o Respond, don’t anticipate, bottlenecks – this is a promarket (antiregulation) position

- “Unregulation of the Internet”o 30 years of deregulation of data networks. o “enhanced services” (not “basic”) should be deregulatedo Business services should be deregulated sooner rather than later. Economics of competition

typically work better in business markets because they are heavy users of telecommunications and have higher requirements.

Reduced interest in consumer protectiono Advanced services

Networks not built yet so regulation is Difficult Destructive (of investment incentives) Less in demand (no gains to transfer)

- Confusiono But…gov’t also intervened to regulate “open access” or “common carriage” for telephone

networkso Oxman glides over distinctions

Deregulation Competition

o Also misses exclusionary policies of “early internet” where commercial traffic was banned. - Internet Backbone

o Economides (2004) describes Network interconnection points (NAPs) Private interconnection

o Traffic exchange via private contracts – networks interconnect per these Ks. o No interconnection mandateso Two tyhpes of interconnection deals

Payment for transift Peering (bill and keep) – interconnect two networks they assume will have the same

traffic. Rather than metering, they don’t measure the traffic they just assume it is comparable.

- Peering



o “free” transit for each networko Works when roughly same traffic and roughly same costs

Costs largely determined by network scope; o Eliminates transaction costs of metering, etc.

- End userso Subscribe to ISPs – most internet access is through enterprises.

- Goldsmith/Wuo Utopianism and the Internet

Internet changes everything “verbs, not nouns” –

o Governments powerless Internet borderless – private actors can escape government rules – Internet utopians

talk about the long arm of the law failing to reach the internet. o Standard controls alive and optimal –

National standards are a good thing on net- Utopianism – Internet Changes All

o John Perry Barlow: “the future will win; there will be no property in cyberspace” – this is dumb – the future always wins. There seems to be a lot of property as well.

o Justice Stevens: the internet “constitutes a unique medium…located in no particular geographical location but available to anyone, anywhere in the world.”

- Internet Borderlesso David Clark: “We reject kings, presidents and voting. We believe in rough consensus and

running code.” – not regulated except by the engineers. o President Clinton: “there’s no question China has been trying to crack down on the internet –

good luck (laughter). That’s sort of like trying to nail Jell-O to the wall (laughter).”- G/W Challenges Internet Myths

o Chinese censorship Cisco routers Google search engines Yahoo, MSN e-mails – Google does not offer mail in China because Yahoo has had

to hand over personal e-mails to the Chinese gov’t where the communications have been suspect. People have ended up in jail as a result of the e-mails turned over by Yahoo.

o Prohibition of Gambling Jurisdictional problems Intermediaries regulated (tradesports.com a UK company accepts bets from VISA,

paypal, etc. which are all companies that can be regulated. If they don’t cooperate, the government

Financial institutions, ISPso Internet is not entirely borderless – you can’t fully get away from geographically distinctions.

This was not technically correct.

Google Book Search: Lichtman; Hal Varian

MISSED CLASS

- Tonight’s Two-fer o Regulation of Broadband networkso Methods for discerning economic effects of regulation

Alternative views of interested parties Market evidence in a natural experiment

- “Open Accesso Runs right into Net Neutrality



o Open access unbundle last-mile connection for competing ISPs. o Net neutrality similar but different way; unbundle last-mile connection via non-

discrimination rules Allow access for rival content, devices, applications – once hooking someone up to

the internet, they have to provide access to other things even if they are not provided by the last-mile provider. Goal is similar but means are different.

- Argument for NN premised on Open Access being eliminatedo NN is needed because “last August, George Bush’s FCC exempted telcos that provide internet

connections from open-access restrictions, dealing a blow to both entrepreneurship and political discourse.”

- Residential Broadbando Cable broadband a “closed platform”o DSL has traditionally been an “open platform” over telco lines. The lines are regulated and

you are required to share. Access fees are regulated. o But then deregulated: obligations of telco networks were reduced.

Feb 2003 – end of line-sharing; Aug 2005

- FCC ends “line sharing” (2/20/03)o dCLECs (data competitive local exchange carrier) pay full cost of local loop; ended policy

where access could be a very low price for dCLECs which would raise the rates of using the LECs local loops.

o “High Speed Service May Cost More” – e.g. if you raise the wholesale cost, then that will drive out competition and Verizon, etc. will raise their rates.

- Full DSL Deregulation (8/5/05o Full deregulation – response to Brand X case. Court upheld FCCs determination that it was

an information service not subject to state and local regulation. Brand X decided the issue of open-access for cable modems. There had never been regulation of cable modems, but there was a threat of it. Supreme Court says the FCC is right. FCC says that they won’t unbundle DSL because it clearly competes with cable. Cable had more than 50% of the market. It was the dominant provider, unregulated. Cable modems were not regulated. DSL was regulated. FCC regulates partially in 2003 and then gets rid of it entirely in 2005.

o “I hope next year the commission will put its money where its mouth is to see if the assumptions yield the results. And if it doesn’t, I hope it will admit that and take appropriate action. I’ll be keeping tabs.

- Arguments for 2005 Deregulationo The benefits of this ruling will ripple across our communities by encouraging greater

investment in and a wider rollout of broadband networks. Phone companies are in favor of deregulation. It gives them more property rights. They can always rent out their lines if they want to.

- Argument Against 2005 Deregulationo “This is a bad day for the internet. I think it means higher prices and less competition and

threatens the growth of the Internet.” – media. - Reliable Evidence?

o The interests – carriers directly regulated want no regulation. They are going to argue their interests which aren’t necessarily in the public interest.

o Regulators – are on both sides of this. o Consumer Advocates – flat out prediction here.

- Empirical Testo Natural Experimento Regime switcho Observe change in reg/dereg service

Output Price Subscribership



o Control Group Unregulated service. Open access rules do not help DSL NY Times headline was ridiculous. Prices went down, and you got much higher

growth rates. - “Keeping Tabs”- Econometrics Tests

o Model 1: DSL sub growth a function of Time trend Same period CM sub growth Deregulation (dummy for post 1Q2003) OR deregulation (dummy for post 2Q2005)

o Model 2: DSL sub growth a function of Time trend Same period Canadian

- Resultso Model 1:

Positive for “line sharing” deregulation Positive for “information services” deregulation

o Model 2 Positive for “line sharing” dereg

- Nominal Prices – w/ Caveats- After the fact, we know the answer. Who were the accurate predictors?

o Manufacturers Intel: Broad-based equipment maker. The broadband market is either going to

expand or not. Intel wants it to expand so there is a broader market for chips. Intel has very strongly opposed open-access rules. Dr. Grove stated that Intel’s goal is to make true broadband widespread and affordable. Intel believes that incumbent companies should not have to unbundled new fiber and DSL equipment as long as they are required to make their existing ‘copper’ lines…” Dr. Groves

Alcatel: switches, routers, etc. – aggressive unbundling and pricing rules can create problems…”

Intel/Nortel: “Intel and Nortel say broadband spending may get a boost from Mon’s Brand X SC ruling…”

- The Rent Seeking Gameo Cable Opponent DSL argues regulate DSLo ILECs Opponent Cable argues to regulate cable. o CLECs Opponent Cable regulate cableo CLECs Opponent ILECs regulate ILECso Note: every sector pro-deregulation (for itself).

- Net Neutralityo “open access” applied at User interface

Open access was imposed at the ISP interfaceo “Allowing broadband providers to segment their IP offerings and reserve huge amounts of

bandwidth for their own services will not give consumers the broadband internet our country and economy need.” Vint Cerf (2005).

- BitTorrent: “I most definitely do not want the internet to become like television where there’s actual censorship…however, it is very difficult to actually create network neutrality laws which don’t result in an absurdity like making it so that ISPs can’t drop spam or stop…(hacker) attacks” A rational set of bandwidth usage might be needed. Neutral rules are okay, but what kind of neutrality?

MISSED A BUNCH OF CLASS

- Wireless & spectrumo Property rights suggested by Coaseo Market allocation of spectrum rather than administrative or public interest allocation of

spectrum.



o Coordinate activities with investors. This is why you need property rights to get value. If everything is unlicensed, investment in cellular networks goes down. Without the security of protections against interruptions, etc. the investment goes away.

- Open access is not good. You have three choiceso Administrative allocation (gov’t property)o State property

Three ways to allocate: 1) you create all the rules. Assigning responsibility for creating rules for resource appropriation. Who gets to use the resource. This is decided by the owner. 2) the government/regulators decide how to create new rights. They can go to governance that say there will be rules about how to use spectrum. There is licensed and unlicensed. They tell you how you can use radio spectrum. You have to transmit at a certain power, here is your business model, etc. You put these restrictions on people in order to reduce interference. In unlicensed, you could do the same thing without regulating the license. Here you regulate the equipment they use. The restriction is the power limit (remote controls, etc. all use unlicensed but it is regulated). In some countries there is regulation of the business model. To restrict crowding, they say you can’t do WiFi for money. Some of them are business restrictions in other countries. You control by governance the unlicensed use. You say what they can do and what they can’t do. They can only use this equipment. They can only use it for noncommercial use, etc.

Traditional license

Liberal license Privatization – you delegate resource use and regulation to another party.

Coase argument is that the regulators should go more towards exclusion. More towards liberal license to identify their own rules. If it is efficient to have unlicensed bands, set it up like a park. Anyone can use it if they pay $5 to enter.

Unlicensed- Tragedy of the Commons

o Some say there isn’t one because there isn’t any congestion and you can fix the congestion. o Tragedy of Open Access (G. Hardin’s oops) – corporations are a commons. Therefore, it is

really a tragedy of open access. There is no residual owner to say this is the optimal amount of grazing, spectrum use, or resource allocation. The actual tragedy is suboptimal valuation. You can have great rules that stop overgrazing but wipe out any value. That is a tragedy as well.

Yogi Berra – “nobody goes to that restaurant anymore because it is too crowded”o Private property

4/17/2007


Date post:	14-Jan-2015
Category:	Documents
Upload:	catharine24
View:	526 times
Download:	0 times

Notes

Documents