INTRODUCTION
The Internet is the fastest growing communications technology in history, far outpacing the adoption rates seen for the telephone, or even television. Today it already represents a major publishing outlet for news, information and many forms of entertainment. In the future it will reach millions of home users as well as businesses and it will continue to influence and profoundly impact not only the lives and lifestyles of an unprecedented number of people but also the paradigms used for doing local as well as global business. Therefore it is crucial that the Internet access and services markets develop in ways that are most beneficial to the consuming public. Consumers are best served when prices are affordable and competitive, and when there are a variety of quality products and/or services from which to choose that are readily available. Currently the market for providing broadband access technology, which is the next generation of Internet access technology, is developing quickly, and is wildly dynamic in terms of mergers and legal activities. However, it is unclear if the industry's current structure and competitive forces will provide and serve the public in the most beneficial manner as described above.
Currently, there are two primary methods competing to provide broadband Internet access to the home Digital Subscriber Line (DSL) technology, offered by the telecommunications industry, and the Internet-over-cable technology, offered by the cable television industry. However, currently the two industries operate under very different conditions due to the Telecommunication Act of 1996, which forced local telecom monopolies to open their infrastructures to competition but, to date, gave no such directive to the cable industry. However, that does not mean the cable industry will not be further scrutinized in the near future to determine whether or not the Telecommunications Act applies to it as well. In fact, the cable industry today very much resembles the telecom industry in the days leading up to the Telecom Act. Like the local telcos pre 1996, today's local cable operators enjoy regional monopolies and control infrastructure over the last mile of cable into homes and businesses. Also like the pre-Act telcos, the cable companies are facing pressures from consumer groups and potential competitors, especially Internet Service Providers, such as AOL, to provide access to their cable infrastructure. The concern is that the local cable monopolies, if not forced to open up, will control all aspects of last mile Internet cable access to the home in the near future. Of course, these potential competitors, who are currently locked out of the cable access industry, would gain substantially if the cable companies were forced to allow them access to their cable network. The question is whether the public would also gain.
In 1996, Congress passed the Telecommunications Act to resolve many of these issues for the telecom industry. Should Congress mandate something similar for the cable industry? To decide which course of action should be taken with the cable industry, one should consider the impact the Telecommunications Act had on the telecom industry since there is a remarkable parallelism between the pre-Act telecom industry of yesterday and the cable industry of today.
Consequently we will begin our study with an analysis of the telecommunications industry and the Telecommunications Act of 1996. We will then examine the cable industry, its similarities to the telecom industry, and the current dynamics that will influence the future decisions. Then, finally, we synthesize these analyses by introducing an economic model as a basis for comparison between the two industries.
TELELPHONE INDUSTRY HISTORY
Before A.T.&T. was dismantled, the telecommunications industry was basically a heavily regulated monopoly that owned, operated, and maintained the entire telecommunications infrastructure. The government forced A.T.&T. to split itself up into several regional "baby bells," also known as Regional Bell Operating Companies (RBOCs), and one long-distance carrier. The RBOCs inherited the infrastructure that connects the homes with the neighborhood central offices, and as a result controlled the local calling market. Competition appeared in the long distance market with the emergence of Sprint and MCI, but since the RBOCs controlled the last mile of infrastructure, they were able to maintain local monopolies. As one might expect from such an arrangement, the RBOCs, in the absence of competition, were slow moving organizations with little market driven innovation and a history of questionable customer service.
Meanwhile, from 1989 to 1996, the market for Internet access from the home doubled every year. At this time, as is still the case today, the majority of home users were using dial-up modems connected to a standard telephone line to access the Internet. Because the RBOCs owned this infrastructure, they were positioned to control Internet access just as they had been controlling local telephone access. At this point they could dictate how, through whom, and at what price customers would access the Internet.
In an attempt to de-monopolize the RBOCs, in 1996 Congress passed the Telecommunications Act, the first major overhaul of the telecommunication laws in almost 62 years. The goal of the Telecommunications Act was to allow free market entry into any communications business in order to promote competition, reduce regulation, secure lower prices and higher quality services, and to encourage the rapid deployment of new telecommunications technologies. In particular, the Act addressed the critical dearth of local competition by directing the regional telecommunications monopolies to open their infrastructure and unbundle their services for resale to competitors at, as the Act stipulated, "reasonable prices."
The Impact of the Telecommunications Act
After the Telecommunications Act was passed, new competitors began entering the market. These new entrants, who became known as Competitive Local Exchange Carriers (CLECs), leased infrastructure from the RBOCs, who, post Act, came to be known as Incumbent Local Exchange Carriers, or ILECs. The CLECs used the infrastructure they leased from the ILECs to provide Local Exchange Carrier (LEC) services to telecommunications consumers, often at cheaper rates than those offered by the ILECs. Examples of CLECs include all of the 10-123 prefix companies that currently advertise so extensively.
During the time after the passage of the Telecom Act, demand for Internet access continued to gain momentum. The ILECs met this burgeoning demand by providing high speed, expensive solutions -- TI, T3 and ISDN -- at high margins to businesses, while home users continued to access the Internet via the slower dial-up modems. Asynchronous Digital Subscriber Line (ADSL) technology was available at this time and offered comparable downstream speeds at lower costs over the same infrastructure as TI and ISDN, and therefore would have better met the needs of the ILECs' home access customers. However, the ILECs chose not to deploy ADSL for fear that it would erode the high margins they received from the aforementioned business solutions.
Some of the CLECs took advantage of this reticence on the part of the ILECs and began implementing ADSL services to homes and businesses. However, in order to implement ADSL, the CLECs had to aggressively pursue their rights under the Telecom Act to collocate equipment in the ILECs' central offices. The CLECs that went this route and began to collocate their own infrastructure alongside that of the ILECs emerged as a new breed of competitor, called data CLECs (also known as "facilities based competitors"). Because of the collocation provision, data CLECs only had to rely on the ILECs infrastructure for access to the last mile, which, as discussed above, was also opened by the Telecom Act. Therefore, companies today, such as Covad Communications Company, which is a data CLEC, are deploying ADSL and are effectively competing with the ILEC’s in the home and business Internet access market. Rather than sit still and watch their share of the broadband Internet access market erode from under them, the ILECs have responded with uncharacteristic speed by offering ADSL themselves. Here, then, is one clear and successful instance in which the Telecommunications Act was instrumental in producing its intended result; because the Act fostered competition, the once complacent, lethargic ILECs were spurred into action to the benefit of their consumers.
Thus far we have concentrated on those companies supplying the underlying infrastructure that provides the last mile connectivity to the end users. But this is not the whole picture. We have yet to discuss Internet Service Providers (ISPs), i.e. the well-known firms such as AOL, Worldnet, Compuserve, etc., that sit between the local infrastructure providers and the Internet backbone. ISPs bundle content and other value-added services such as data caching, web servers, and email facilities and offer them to their customers at competitive prices. ISPs have traditionally been customers of the ILECs, since there are several business telephone lines installed at every ISP facility. The dial-up modems of the ISP customers connect to the ISP’s modems via these telephones lines. The ISP modems then connect to switches that route the signals to the Internet backbone.
ISPs are important because they control so much of how the end user experiences the Internet. For instance the ISPs control what information content their customers see first, and how the billing is done. Before the Telecom Act, RBOCs were positioned to control many of these functions by bundling their own ISP services with their infrastructure. The Telecom Act changed this by mandating that the local infrastructure companies, the ILECs and data CLECs, offer a selection of ISPs to their customers. Furthermore, the ILECs and data CLECs were forbidden to use their infrastructure positions to give their own ISPs an advantage over competing ISPs. This openly competitive environment spurred the entry into the market of numerous ISPs, who now offer ADSL service through ILECs and data CLECs, but who still rely on customers dialing up through modems for the vast majority of their business.
Complications Resulting from the Telecom Act
Despite promoting the desired competition, the Telecommunications Act also introduced a plethora of complications and unwelcome litigation for all the players. As mentioned above, the Act mandates that the ILECs lease its facilities to the CLECs at "reasonable rates," regardless of what facilities are being leased. However, what is a "reasonable lease rate" is not clearly spelled out by the Telecommunications Act, and therefore not all parties agree on the interpretation. To solve the resulting disputes and grievances the Act allows for statewide commissions to hear the matter and arbitrate some kind of fair judgment.
However, the fact that every state has different standards and pricing rules produces a regulatory and litigation nightmare. CLECs have to first register for and receive permission to become a Local Exchange provider from the Public Utility Commission in each state it wishes to enter. They then have to sign Interconnection Agreements with the ILECs in that state. If they have grievances with the "reasonable rates" the local ILECs make available to them, their only recourse is to go to the state commission and hope for a better "reasonable rate". For CLECs trying to do business in more than one state, this procedure is prohibitively awkward, time consuming, and costly.
On January 25, 1999, the U.S. Supreme Court remedied this problem in part by finally ruling that the FCC, and not the states, has primary and plenary power to interpret Sections 251 and 252 of the Telecommunications Act and for setting a ceiling for service prices. By keeping the price authority at the FCC, it is hoped that competition will be facilitated as ILEC competitors (CLECs) will not have to deal with 50 different state policies and legal processes, thus effectively lowering the barriers to entry to the industry for these CLECs.
CABLE TELEVISION AND BROADBAND INTERNET ACCESSES With the recent explosion in popularity of the Internet, and the
subsequent demand for greater bandwidth access by home users, there is
a potential fortune to be made by whomever becomes the dominant
provider of broadband access. Seeing this, CATV providers are eagerly
making and implementing their plans to enter the broadband market. A few large companies known as Multiple System Operators (MSOs)
dominate the cable television market in the United States. The top 5
MSOs, TCI, Time Warner, MediaOne, Comcast, and Cox, account for over
half the CATV market.
[1]
Each of these MSOs owns many individual, regional cable systems, and
typically the MSO enjoys a monopoly in each region it serves. Though cable operators already have in place extensive distribution
networks that include coaxial cables to each subscriber’s home, most
cable operators must make substantial, and costly, improvements to
their networks before they can offer internet access in a given
region. CATV Networks Historically, a typical cable television network serves a city
or region with a size ranging from a few thousand to a few hundred
thousand homes. The network has a tree topology, where the cable
subscribers can be thought of as the leaves of the tree, and a central
station, called the head end (H/E), is the root of the tree. The head
end combines television signals from over-the-air and satellite feeds
and transmits them across trunk lines that go out to different
neighborhoods or sub-regions of the service area. A system of branch
lines at the end of each trunk distributes the signals to the
individual subscribers in each sub-region, typically 500 to 5000
customers. In the past cable television networks were constructed using
coaxial cable, but now most cable operators have replaced the trunk
lines emanating from the H/E with digital fiber-optic cables because
they reduce noise on the network and increase capacity (i.e. allowing
operators to offer more channels.) The resulting network is often
called a hybrid fiber coax (HFC) network. Downstream video signals
usually occupy the range between about 50 MHz and 800 MHz of bandwidth,
leaving a significant portion of the spectrum below 50 MHz, i.e. 5 MHz
to 42 MHz, free. Modern HFC networks use some of this bandwidth
to allow for upstream communications from the subscriber's home to the
head end, typically for requesting pay-per-view programming. Such
networks are called two-way HFC networks. Converting CATV Networks into Data Networks To offer data services over a cable television network,
operators must first upgrade the network in an entire region to
two-way HFC. This can be a costly process of changing
one way amplifiers throughout the network into two-way
amplifiers. Next, the cable operator must make a large investment in
equipment in the head end station, and allocate one television channel
for downstream communication. The typical cost to make these
improvements range from $200 - $250 per home,
[2]
though a few cable operators have avoided some infrastructure
investment to two way HFC by having users send all
their upstream data via regular telephone modems in a system known as
Telco-Return. Once these infrastructure investments are made, the only additional
cost for adding an additional subscriber is the cable modem at the
subscriber’s end (typically about $350.) This is in contrast to
ADSL, where with each new subscriber the data CLEC must purchase a new
line card at the central office in addition to the DSL modem that must
be purchased on the subscriber’s end. Cable modem networks have a
smaller marginal cost for an additional subscriber, but a greater
up-front fixed cost from the region-wide network improvements that are
necessary before anyone can be offered cable access service. The bandwidth of a cable modem network is asymmetric. The
downstream bandwidth varies between 27 Mbps and 40 Mbps, while the
upstream bandwidth varies between 500kbps and 10 Mbps (depending on
the amount of spectrum allocated for the service, and the noisiness of
the network.) Furthermore, the users at the end of each fiber trunk
line, which can number between 500 - 5000 subscribers, share both the
upstream and downstream bandwidth. In contrast, ADSL bandwidth is not
shared among multiple subscribers.
This means that cable access speeds are subject to degradation whereas
service through DSL is not. However the resource sharing
inherent with cable modem technology means that it enjoys better
economies of scale than DSL, and potentially can serve a large
number of people more efficiently. One study says that 225 nodes
connected to a central office with DSL technology would cost
$1300 per connection, whereas the same number of nodes in a cable
system would cost only $450 per connection [7]. Though, the same study
argues that the gap will close as DSL modems drop in
price. Offering Internet Services Offering Internet service requires more than upgrading the
cable network into a data network. It also requires acquiring
connectivity to the Internet backbone as well as Servers, Network management
tools, billing systems, and security systems.[1]
In addition, providing quality service requires having a large amount
of popular, multimedia content cached on servers, ideally right at the
head end, so that users can get the maximum benefit from their
high-speed connections. To provide these services, MSOs almost always
partner with a single company specialized in Internet services for
cable, or more simply Internet Service Provider
(ISP). In several
important cases, these ISPs were actually created by the MSOs as
separate business ventures. The two most important examples are
@Home, and The Road Runner Group.
@ Home
was formed by TCI in May 1995, and received
subsequent equity investments from Comcast Corp., Cox, and
Cablevision. As should be expected, all of the MSOs who have a stake
in @ Home partner with it exclusively when offering
Internet services.
This gives @ Home a potential installed base of 53
million homes, more than half the cable homes in North America.
[2]
The Road Runner Group
was formed by Time Warner as a separate business. Subsequently, the
MSO, MediaOne, formed an alliance with The Road Runner
Group, and technology giants Microsoft and Compaq invested a
combined $425 million in the venture. Today, subscribers of MediaOne
and Time Warner must use The Road Runner Group as their
ISP.
INDUSTRY TRENDS AND AN UNCERTAIN FUTURE
Though it is still early in the race between cable and DSL,
cable-modems have a strong early lead. About 20 million homes in North
America have access to cable Internet services, and of those, about
500,000 have subscribed compare to only 25,000 DSL users. Cable
operators are upgrading their networks region by region, and will
spend an estimated $10 billion this year on improvements.
[3]
However, some people are frustrated that cable companies aren’t
making the upgrades faster.
[4]
With cable appearing more and more likely to become the dominant
broadband access technology to the home, other companies are racing to
get a piece of the market. The biggest and most high profile example
of this has been the merger of AT&T and the largest MSO, TCI, in
early 1998. With hopes of becoming a dominant provider of broadband
Internet services, including IP telephony, AT&T acquired TCI for
$48 billion, and is planning to spend $5.7 billion on network
upgrades. In more recent news, AT&T, Comcast, Microsoft, and
America Online are all fighting to purchase another MSO, Media
One. America Online (AOL) may have the most at stake of all the
companies fighting for Media One. America Online is the largest ISP
for users of dial-up modems, but as users switch from dial-up modems
to cable-modems they will have to switch from AOL to the single ISP
offered by their local cable company. Without any new government
regulations, the only way to be sure that a cable company offers AOL
is for AOL to buy the cable company. But AOL, other dial-up ISPs, and consumer groups, are calling for
the government to start forcing cable operators to allow their
subscribers to chose other ISPs. However, in the context of the
AT&T / TCI merger, the FCC has recently decided that it should not
apply "common-carrier" like rules on cable modem networks,
because they feel that the broadband market is in a ‘nascent
stage’ and must evolve free of intervention.
[5]
Yet, the issue is still far from dead. In April, 1999, Stephen Case of
AOL testified before the Senate Commerce Committee calling for cable
operators to open up to other ISPs, and Senator John McCain has said
that he will introduce legislation asking the FCC and the Commerce
Department to offer recommendations. The question is even under scrutiny at the state and local
level. ISPs have been encouraging the introduction of state
legislation which would force cable companies to open to other ISPs,
while the cable industry has been arguing fervently that the states
have no right to regulate them.
[6]
Recently, Multnomah County, OR required that AT&T / TCI open up to
other ISPs besides @ Home or else the county would revoke their
cable franchise rights to the area. AT&T / TCI promptly sued
Multnomah County, and the matter is still before the courts. Clearly, AT&T / TCI and the other major MSOs appreciate the
enormity of the potential market for Internet access over cable, and
will fight to make sure that they share the market with as few other
players as possible. At the same time, the MSOs are spending billions
of dollars to gear up their systems so that more and more people can
have access to the Internet via cable. So far the industry has been
left to grow unfettered by additional regulation, as the FCC has been
weary in bringing cable companies under the fold of the Telecom Act
and treating them as they treat ILECs. Is now the time to put the
breaks on the power of monopoly MSOs, or would we just be interfering
with the growth of an exciting new industry?
ECONOMIC MODELS: A COMPARISON OF THE TWO INDUSTRIES
In deciding which course of action is optimal, we must consider what
is best for the consumer. The objective of both sides will be to
argue that their position will grant the consumer the most value,
which we measure in terms of lower prices, higher quality, variety of
choice, and broader access to broadband Internet access service. In
order to build a sound argument for either side of this debate, we
believe it is important to consider the economic models at play in
both the telecom industry and the cable industry. While there are
certainly numerous differences between these two industries, we
believe the parallels are significant enough for us to make useful
comparisons. As mentioned above, the telecom industry prior to the
Telecom Act of 1996 looked and behaved very similarly to the cable
industry of today, with regional monopolies controlling the last
mile infrastructure, the availability and quality of services,
and the prices being charged. In addition, like the cable consumers of
today, the telecom consumers had but one choice for their local
carrier company. The Telecom Act changed the whole landscape for the
telecom industry by mandating competition for the RBOCs. Now
the question is whether the cable industry should be given similar
legislative directives that would require it to accommodate
independent ISPs and other competitors.
To better understand this question, each industry will be individually
analyzed using a generic cost and revenue model. By using the same
basic model for both industries, the similarities and differences
between the two industries should be readily apparent. Using this
comparative insight, a well understood analysis of alternative answers
to the question posed above should be forthcoming along with a solid
foundation on which to build the accompanying supportive
arguments. The Model
The telecom and cable companies, like all other companies, are
striving to maximize the present value (PV) of their future
cash flows, which comprise revenues as well as both fixed and variable
costs. The ratio between these fixed and variable elements is crucial
to understanding the competitive motivations of firms within each
industry. To illustrate these costs and revenue flows we will utilize
a PV equation that a telecom or cable company might use when
deciding to provide service to a new area.
PV = Q *[Installation Fee + ( (1/R) * (Monthly Fee) * S ) –
Variable Cost] – Upfront Cost Where:
A cable or telecom company that computes a positive PV for introducing broadband services to a new region would most likely proceed to do so. Otherwise, if the computed value of PV is negative, the company most likely would choose to stay out of the market, potentially leaving consumers without access to broadband services. One other crucial point is that the number of subscribers, Q, and the Monthly Fee the service provider will be able to charge cannot be forecasted perfectly. A telecom or cable company would like to see that the PV on a new project would remain positive even if demand in the area turns out to be lower than expected.
The Telecom Industry
In our discussion of the economics of Internet services from the telecom industry we focus on the DSL technology, the most popular and economic broadband technology the telcos offer. DSL technology uses an existing phone line, a DSL modem at the subscriber end, and a line-card at the central office to provide broadband service. Because last mile telephone lines are already in place, there is no major cost for laying these telephone lines, and thus there is no Upfront Cost that is relevant to an ILECs’ future decisions. The remaining portion of the costs associated with DSL, the subscriber modem and the line cards at the central office, are Variable Costs, since these costs to the ILEC scale with the number of subscribers. Because the costs are mostly variable, the ILEC assumes very little risk in offering DSL to new areas. The ILEC only spends money each time a new customer signs up, and thus does not assume the risk of investing a large amount of money into the infrastructure before being able to offer services to a group of potential, would be customers. The risk remains relatively small even when the customers choose an alternative ISP. In this case, the ILEC receives the Installation Fee along with a share, S, of the monthly fee that it negotiates with that ISP. Therefore in the case of the Telecom Industry, a positive PV is relatively easy to achieve because of the practically non-existent Upfront Costs.
The Cable Industry
The economics of providing cable Internet access are very different from the economics of providing DSL access. To offer service to a new region, a cable company must pay a large Upfront Cost to install the head-end gear and to upgrade the new region’s existing cable infrastructure for two-way communication. If cable companies anticipate that the revenue stream from the Monthly Fees will cover the Upfront Cost, they will have an incentive to build infrastructure in new areas. If they are uncertain that the revenue stream will be large enough to cover their Upfront Costs, they most likely will choose not to introduce cable access service in that region since that would imply a negative PV.
At present, with no ISP competition, MSOs act as local monopolies, and being free of competition they can thus maximize their profits. Their monopoly power aids MSOs in recovering their Upfront Costs, but it arguably can also result in a less than beneficial situation for its customers. For example, Figure 1 below, depicts an MSO's cost and revenue situation for one single built-out region. (See above for details on the logistics of building the cable Internet access infrastructure.) The concave, downward sloping Average Total Cost (ATC) curve reflects the large upfront fixed cost of installing a head-end station and then the relatively small marginal cost of connecting each additional user in the built-out region. In fact, we assume that marginal costs are flat, as there seems to be very little, if any, diminishing returns to scale once a head-end station that supplies sufficient bandwidth has been installed. As a monopoly, an MSO will set price and quantity at levels that maximize its profits, represented by the shaded area in the graph (which is a rectangle whose area has been maximized). In a more competitive situation, the quantity installed would be larger and the profit smaller, which would yield a larger consumer surplus area and a rectangle whose area would be smaller. Thus the profit maximizing equilibrium for the monopoly is reached without regard for the customer by extracting from the consumer surplus that would exist under competitive situations. Therefore the result for the consumers is a higher price and a lower availability of service than would exist in a more competitive market.
Figure 1. Given the assumptions reflected in the graph which are hypothetical values used here for illustrative purposes only, a monopolist would produce at a price of $44.29, yielding a quantity of 6,076, for a penetration of only 60%, but for a monthly revenue of $203,240. By keeping prices high while limiting quantity, the MSO would be exercising its monopoly power, which anti-trust economist Frank Fisher describes as "the ability to act in an unconstrained way." In a more competitive market a reasonable price might be $20, which would yield a quantity of 9,200 and a penetration of 92% which indicates wide availability. However, the associated monthly revenue would be only $92,000, which is considerably less than the $203,240. Thus competition created a situation which traded a benefit for the consumer for a loss in profit for the monopoly.
In particular, if cable companies providing Internet services are directed to open their networks to competitive ISPs instead of being allowed to exclusively use @Home or Road Runner, (the ISPs owned by TCI and MediaOne, respectively,) the additional competition would drive the price to the customer down, and at the same time convert some of the supplier surplus into consumer surplus. On the down side, ISP competition might reduce revenues to the cable company enough so that there will no longer be an incentive for the cable company to build infrastructure in new areas. Once again this lowers availability to the consumers, and no one has benefited by the introduction of ISP competition.
A Model for Cable Unbundled from ISPs
By allowing competing ISPs access to the MSOs infrastructure, a two layered industry is created. The underlying layer -- the infrastructure, built by the MSOs -- remains a monopoly, while the top layer -- Internet access service -- would be open to competition.
The unbundling, and the resulting two-layer market, will have important implications for the MSO' revenue models and therefore for their infrastructure building plans.
In this new model, the price paid by the consumer, P, will consist of two components:
The two parties, the MSO and the ISP, would have to negotiate a PMSO that would be beneficial to each as well as to the industry as a whole. The goals of such a negotiation would be as follows:
However the two layers in this model do not have equal negotiating power. The lower layer, the MSO, is a monopoly, while the upper layer, the ISP, is subject to competition. The competition will exert downward pressure on PISP. Meanwhile, the MSO will be attempting to increase PMSO as much as possible. These conflicting forces are likely to generate tension during negotiations and might inhibit the two parties from arriving at a revenue sharing agreement that is most beneficial for the consumers and the industry as a whole.
FINAL REMARKS
Unbundling cable and ISP services might reduce consumer prices, but with such complicated dynamics in this new two-layer model, it is unclear whether cable companies would have the incentive to broaden cable access by building more infrastructure. So, should cable companies be forced to open to ISPs? Arguments can certainly be made either way.
BIBLIOGRAPHY