The visionaries sketched it out, the computer
literati caught onto it, and now it's all over the mainstream media: a high-speed,
fiber optic, new information highway carrying an expanding universe of information
and entertainment into the home and the workplace. Thousands of movies, mail-order
catalogs, newspapers and magazines, educational courses, airline schedules,
and other information databases will be available with a few clicks of a remote
control. Two-way video conferencing will revolutionize business meetings, visits
to the doctor, and heart-to-heart talks.
This vision has been tirelessly promoted for more than a decade now but
with little visible progress toward its realization. Political gridlock has
snarled telephone companies, newspaper publishers, cable television operators,
and other potential players in lengthy and fruitless congressional and court
battles. As a result of this logjam, a justifiable cynicism has developed
to fill the gap between vision and reality.
Meanwhile, pioneers in the on-line worlds of computer-mediated communication
networks, collectively referred to as Cyberspace, have not been sitting idly
by. Employing whatever tools they can find, they are constantly pushing at
the techno-cultural envelope. Life in Cyberspace is often conducted in primitive,
frontier conditions. But at its best, it is more egalitarian than elitist,
more decentralized than hierarchical. It serves individuals and communities,
not mass audiences; cuts across national boundaries; and is extraordinarily
multi-faceted in the purposes to which it is put.
In fact, life in Cyberspace is shaping up just as Thomas Jefferson would
have wanted it: a decentralized democracy, founded on the primacy of individual
liberty and committed to pluralism, equality, and community.
Now, for the first time, something is happening with the major industrial
and political players, too. A broad consensus is emerging on key business,
technological, and political aspects of a National Information Infrastructure
(NII). The creation of high-capacity networks by cable and telephone companies
Yet even as the crucial agreements between business and government are forged,
fundamental questions remain about the implications of the re-wiring of America.
Will we get the openness, freedom, and diversity that represent the true promise
of new information technologies? Or will we end up instead with networks controlled
by mega-corporations, fostering addiction to a new generation of useless electronic
narcotics (glitzy, interactive multimedia successors to Nintendo and MTV),
and encouraging instant gratification through sex and violence?
The emerging consensus between business, government and policy watchdogs has
three key points:
Private, not Public. The private sector, not the government, will build
and operate the NII. The government will fund research, lead the effort to
develop experimental ultrahigh-speed networks, help to promote industry standards,
and protect freedom of speech and rights of privacy. But the principal carriers
of traffic into the home will be telephone companies and cable television
operators, not the government.
A Hybrid Net. Much of the public discussion about the new information highway
has focused on fiber optic technology. But the networks reaching into the
home will actually be hybrids of fiber optic cable and the existing copper
wire and coaxial cable used by telephone and cable television companies. Fiber
optic cables will be used in the major arteries and portions of the distribution
system; copper and coaxial cable will be used in the last hundred yards. To
achieve a broadband network capable of delivering high-quality video, voice,
and data, it is both unnecessary and too expensive to replace the last segment
into the home with fiber optics.
Video Driven. In the near term, the principal business opportunity driving
investment in a broadband network will be the delivery of movies on demand
to the home. Government regulators are now trying to orchestrate an overall
path towards more competition in this emerging market, and the decisions they
make will leave an important imprint on the future evolution of the NII.
This three-pronged consensus fixes the very broad directions of the new information
highway. But it also leaves a wide range of possibilities open, with vastly
different implications for the Jeffersonian vision.
The optimistic view of the next decade is that cable and telephone companies
will be led by a logic of self-interest to build networks consistent with
Jeffersonian ideals, even if they are unaware of or uninterested in the full
range of democratic goals. The key assumption in this optimistic scenario
is that networks permitting the greatest diversity of content and services
video programming, computer software, and whatever else is dreamed up by the
mass media and individual users will create the biggest market and the largest
sustainable business opportunities.
Moreover, if the communications industry is responsive to key ideas about
openness, then the prospects for the optimistic vision are even greater. Openness
means a number of different things access to the highway for all citizens;
control by users over the services they use and when they use them; relatively
easy entry for new firms. If adopted as a guiding principle, openness as exemplified
by the personal computer and such computer networks as the Internet can be
a driver of prosperity and diversity in any new market.
The pessimistic scenario is that costs of access to most services other
than entertainment will be high; content will be supplied by a carefully chosen
set of providers, with most users left in the position of consumers rather
than producers of information; networks will be controlled by a few firms;
and programming will seek the lowest common denominator. The result will be
a population divided by income into information haves and have-nots.
The future lies somewhere between the optimist's and the pessimist's views.
But precisely where depends on policy decisions that will be taken in the
near term. Optimism, combined with vigilance and a commitment to seek government
intervention for redress of private enterprise's failures, is the best prescription
for coping with the NII's uncertain future. But before jumping to recommendations
about public policy, let's first get an understanding of the major players
and the changing technologies.
The Emerging Industry
The big players in NII development are cable and telephone companies, each
driven by increasingly sharp competitive pressures to accelerate broadband
deployment. The regulatory barriers which have kept cable and telephone companies
out of each other's base businesses are surely going to fall in the next decade.
The result is that nobody's market share is protected. If telephone companies
can offer video programming, cable revenue will surely drop. If cable companies
can offer local phone service, the Regional Bell Operating Companies (RBOCs)
will be hit where it hurts.
These pressures have produced increasingly rapid movement across market
boundaries. Earlier this year, the two largest cable operators TCI and Time-Warner
Cable announced major infrastructure upgrades. TCI plans to increase the channel
capacity from 50 to 500 channels through the use of digital compression. Time-Warner
is developing a “Full Service” network in Orlando, Florida, which will offer
cable television, video on demand, and fully interactive capabilities.
The RBOCs are not standing still either. Bell Atlantic, for example, will
be testing an interactive, switched, broadband network in New Jersey beginning
this year to deliver cable television, telephone, and other services over
the same wire.
Finally, the recent acquisition of McCaw Communications by AT&T creates
the first integrated local and long-distance telephone company since the breakup
of the Bell Capital system a decade ago. This positions AT&T to move independently
of the RBOCs in providing new consumer services.
A major target for both cable and telephone companies is an emerging video-on-demand
market, competing with the $12 billion market for video rental. The viability
of the market for video-on-demand has become an article of faith, based on
extrapolation from a fundamental law of nature: no matter how many copies
of a hot movie a video store has, there's never one left on the shelves when
you get there. With video-on-demand this will never happen, as individual
digital realizations of a stored film image can be generated as needed. Add
the convenience of skipping the trip to the store, and being able to make
last-minute decisions about what to watch, and there is a compelling case
for the inevitability of movies on demand.
But major market opportunities are not confined to video-on-demand. Cable
companies want to compete for a share of local telephone business (an $86
billion market) by offering a new generation of cordless phones that work
through a cable set-top converter. Telephone companies want to supply basic
cable television service, and a recent Federal District Court ruling overturning
restrictions on RBOC provision of local cable television service points them
in this direction. And everyone is hoping major new markets for interactive
shop-at-home services will open up.
A final piece of the competitive picture is that the computer industry has
now joined the party. In April of this year, Microsoft and General Instruments
announced the intention to work together to create the intelligent set-top
converter required by the new network. Really a computer in disguise, this
next generation cable box will contain an Intel 386 (or higher) microprocessor
and multiple megabytes of ROM and RAM.
How all this maneuvering will pan out is of course uncertain. But one thing
seems clear that cable companies are better positioned than telephone companies
to take advantage of emerging opportunities. RBOCs themselves are pursuing
very different strategies, indicating a lack of consensus on the winning broadband
formula. Some may not survive the transition to competition in local telephone
service others, cognizant of the threat to their installed bases, have bought
into cable franchises. But telephone companies are clearly playing catch-up
when it comes to providing entertainment services over their facilities. In
addition to the technological risks they face (which are shared by cable)
they are hindered by a regulatory environment which still contains substantial
prohibitions. Moreover, they lack the kind of entrepreneurial culture that
encourages rapid strategic adjustments in dynamic environments, big bets,
and the will to see those bets through all of which puts them at a disadvantage
relative to cable operators.
The technological bases of NII have been shifting as fast as the industry
players. Infrastructure rhetoric of the 1980s called for end-to-end fiber,
so much so that politicians and press alike have mistakenly identified fiber
optics and broadband networks. But the current industry consensus reflecting
new technologies for information compression, storage, and transmission is
that networks will join fiber optics to existing coaxial cable or copper telephone
wire. Such hybrid networks have the potential to deliver a full range of interactive
services at a fraction of the cost of fiber-to-the-home.
To accommodate an interactive network, cable television's physical infrastructure
will need to be substantially transformed. Today's separate cable systems
will be joined via regional hubs, which, in turn, will be interconnected to
form a national network. High-capacity video file servers capable of storing
thousands of hours of programming will be attached at the regional level.
Existing cable systems will be replaced by fiber optic cable that will reach
from the head end of the cable system to a node serving a group of 200-1500
homes. The last segment of the network from the neighborhood node to individual
homes will use existing coaxial cable.
This network architecture will be capable of carrying hundreds of channels
of programming into every home, as well as custom programs selected by individual
households. Initially almost all of the capacity will be used to bring signals
into the home, with very little being used to carry traffic back “upstream,”
from one user to another. But with proper upgrades, a hybrid network also
will be capable of accepting and switching telephone traffic. Over time, the
network can be expanded to support full two-way communication applications
like video telephony.
Broadband networks for telephone systems are rapidly converging on delivery
of the same set of services as the emerging cable network. Here again, deployment
has been aided by the development of a new transmission method called “Assymetric
Digital Subscriber Line” or ADSL that makes it possible to use existing copper
wire to deliver video to homes. ADSL is not yet a perfect substitute for cable:
it is limited to sending one channel at a time and will not serve households
in which more than one television is turned on. But these limits are temporary.
Within a year or two ADSL will be able to carry live news and sports and serve
A Policy Prescription: Openness
Let's assume all the industry strategies and all the technology shake out
in unpredictable ways by the end of the decade. While the precise outcome
of battles between cable and telephone superpowers will make a difference
to the consumer, more important still are the basic design principles both
in technical architecture and in public policy under which the winning entry
or entries operate. And the key design idea for the Jeffersonian path is openness.
Openness is an issue in every aspect of the network. A network is either
open or closed with respect to who may have access to it, who may supply content,
who determines its specific uses, and how its architecture is determined (who
can provide the equipment, how interfaces and standards are determined, and
whether the technical details are public or private).
The Internet is the paradigm of an open network. It is an interactive medium
based on two-way communications, where people who provide their own equipment
can fluidly shift between positions of listener and speaker. The public telephone
network is substantially closed in its architecture, but, by virtue of being
common carriers, telcos are required to be open in access, content, and use.
Cable systems, on the other hand, have no such obligation, and exercise very
tight control over both content and use.
The Jeffersonian option requires a commitment to openness in all of its
dimensions. And it requires that commitment right from the start: it is much
easier to build in openness into networks then to add it after the fact. Building
an open network will require that policy makers and businesses ask themselves
some fundamental questions before committing to any one path.
Who has access to the network? Is it affordable? Many basic human services
in health care and social welfare could be transacted over a ubiquitous voice,
data, and video network far more easily than by requiring the elderly, the
infirm, and young mothers with small children to take public transportation
to municipal, state, and federal office buildings. But without guarantees
of an affordable connection, the network will further stratify society, not
bring it together.
Everyone has the right to telephone service, and local telephone service
is inexpensively priced. But universal service has been achieved through a
system of regulation which is widely criticized and in the process of being
dismantled. Patricia Eckert, Chair of the California Public Utilities Commission,
speaks of the co-dependency of regulators and the regulated. The cure for
this addiction is more competition. But no one can promise that competition
alone will ensure universal, affordable service. Universal service is the
baby which must not be thrown out with the bathwater of a dysfunctional regulatory
system. In truth, no one knows how to do this yet. It is therefore imperative
that, in the public policy debate about broadband networks and increasing
competition in local phone and cable service, a universal right to service
be given priority.
Who can put content onto the system? In the “video dial-tone” model developed
by the Federal Communications Commission (FCC) for telephone companies (which
has already been approved), telcos will be allowed to make content available
through their own content servers. But they are also required to allow third
parties to attach their own video servers to the network, and to charge on
a non-discriminatory basis for transport of the content to subscribers who
request it. In fact, the FCC has gone so far as to specify that the top-level
menu or “gateway” which appears on the user's screen points to the third-party
In this respect, the terms of providing content under video dial-tone would
generally follow the practice on the Internet there, anyone who wishes to
make information available is free to put up an FTP, Gopher, or WAIS server,
to which any other Internet user can connect.
Ideally, any user of any network should be able to originate programming
from the home. To do this requires a network with sufficient upstream bandwidth
to carry at least one compressed video signal (optimally, a minimum of 1.5
megabits/second). Either the basic network itself should have this capability,
or, if this is uneconomical, it should be an option which is always available
as an add-on. In either case, it should be as easy to provide a service as
it is to use one.
In the on-line world, open and closed models exist. Commercial services
like Prodigy and, to a lesser extent, Compuserve and America Online, tightly
control who can put information onto the service. At the same time, the existence
of 45,000 individually-controlled bulletin boards testifies to the fact that
a lot of people want to run information services. Networks must be designed
to support sufficient upstream bandwidth to allow for maximum flexibility
in point of origin.
The model for cable television has been completely different. While cable
operators are required to carry most local, over-the-air broadcasts and a
small number of public access channels, they are otherwise free to do what
they want. An industry structure in which major cable operators are also owners
of programming creates a further incentive to restrict content. Time-Warner
and TCI, for example, both hold major shares of CNN. They might just think
twice before agreeing to carry another 24-hour news service.
The cable industry would need to reinvent itself substantially if it is
to open up in terms of content. The federal government has recently given
cable a big incentive for such reinvention: the 1992 reregulation act put
a tight cap on revenue growth from basic cable services. So growth will require
developing new services. These new services might eventually include the use
of cable as a carrier for unrestricted third-party provision of content from
alternatives to CNN to home videos. In this scenario, enlightened self-interest
will eventually drive cable to more openness in content. Failing that, the
government must develop a new type of legal regime which obligates a carrier
not to discriminate in content, but does not impose heavy, common carrier-style
Will novel uses of the network be allowed to develop? In Massachusetts,
Digital Equipment Corporation and Continental Cablevision conducted an experiment
using idle capacity on an in-place cable system as a ten megabit/second local
area network. Technically, cable companies could add a low-cost data service
option for access to the Internet and other computer communication services.
But cable system operators, who think of themselves as vehicles for the delivery
of mass-market entertainment, are likely to overlook the opportunity here
because it is so far outside their usual province.
New and potentially revolutionary services such as these should not be overlooked
or squeezed out of the market. If users have more control over determining
the uses of the networks these new opportunities are more likely to get the
early nurturance they need to turn into big, recognizable business opportunities.
To accomplish this, we should consider setting aside experimental frequencies
on broadband networks for developing significant new uses. This approach is
gaining favor in the wireless world, where the FCC is now granting valuable
Pioneer's Preferences for such developments, and it should be considered for
adoption in broadband as well.
Will system specifications and interfaces be publicly available and defined
in an open process? Both telephone and cable companies will develop interface
specifications for video servers. Will these be publicly available? Or will
the standards be proprietary and controlled by the manufacturer? Computer
networks adhere to the open model, while, historically, telephone company
central office equipment and cable head-end systems have been quite closed.
If specifications can be obtained only through special pleadings, the garage-and-attic
segment of developers who create the most radical innovations will be overlooked
or even excluded. This must not be allowed to happen.
500 Channels and Nothing On?
Let's assume that in another ten years we all have powerful, user-friendly,
innovative gizmos in our living rooms. The basic question remains: Will there
be anything worth watching? Can we expect a real diversity of sources and
uses of information, or simply an emptier wasteland?
The advantages of diversity seem clear enough, particularly from the Jeffersonian
angle. Increased diversity will permit more raw, unfiltered sources of information
to reach the people. During the last presidential campaign, more people began
to turn off the blow-dried network TV anchors and turn on C-SPAN. Professional
pundits were shocked by the size of the audience for the debates and town
hall meetings. Direct participation in the political process, an idea fueled
by Ross Perot and supported by Bill Clinton, would be well served by open
platform for video. Coupled with the interactive capabilities of computer
conferencing systems, the dialogue might replace the monologue as the staple
of political discourse.
Some argue that diversity will lead to social fragmentation. Others say
that most content will be junk. But print, the medium of the greatest diversity,
reflects our culture rather than fragmenting it. And noise is the price we
pay for signal. In fact, without junk, there is less of a chance for real
quality to emerge. Today's noise is tomorrow's signal.
Assuming, then, that diversity is an important policy goal, the question
is how to achieve it. Here, we can learn an important lesson from the history
of cable television.
One of the arguments used in favor of cable was that it would increase both
the number of channels and the diversity of programming. The prediction about
numbers was certainly right: the average cable system carries dozens of channels
which aren't available over the air. But real diversity is qualitative as
well as quantitative, and as cable shows we do not get qualitative diversity
simply by increasing the number of channels. If systems are controlled by
a single operator per market then no matter how many channels there are the
desire to seek the highest economic return per channel drives programmers
to seek the largest possible audience and the lowest common denominator. So
cable channels are now filled with 1950s reruns, shopping channels selling
overstocked merchandise, and other commercial detritus. And as the number
of cable channels per system continues to rise, we are seeing more and more
niche cable services comedy, science fiction, and cartoon channels. It's all
the same old television pie, cut into new, thinner slices.
Diversity, then, is not just a matter of numbers. Instead, it raises fundamental
issues of architectural design, political control, and business strategy.
As a matter of technology, diversity requires a switched architecture in
which some channels don't carry fixed content at all. Rather, a given channel
serving the home should be able to carry any particular program content in
the same way that the telephone in the home today can be connected to any
other telephone in the public switched telephone network.
Above all, though, diversity is now a matter of distribution. Until recently,
there were bottlenecks of expensive resources at every stage in the video
chain particularly in production. But thanks to advances in video cameras
and desktop video, we're down to a single chokepoint: distribution. Without
an inexpensive, accessible, affordable infrastructure for the distribution
of video material, the diversity of useful video material will be limited,
and the potential of the medium will not be reached.
It should be a national goal to create an infrastructure that removes this
bottleneck and permits the unencumbered distribution of video programming
of all kinds. A worthy model is the infrastructure which supports the distribution
of print material the postal system, roads and highways for the transportation
of bulk print matter, etc.. This infrastructure includes a taxpayer-supported
network of roads and highways for the physical movement of printed material,
and subsidized postal rates for publications. The resulting richness and diversity
in print media can serve as a role model.
A national video infrastructure may not require taxpayer subsidy. Entertainment
and other premium services alone will justify the broadband investment, according
to the cable and telephone companies that are building them. But it does mean
policies that foster a shift from an ecology of a small number of instances
of relatively high quality to a large number of instances of wildly varying
quality. Over time, as a new generation of professional and non-professional
users learn to master the tools, and as the tools themselves improve, the
average quality of production will increase.
The critical public choice regarding the information highway is this: If industry
builds it, how happy are we likely to be with the result? To what extent will
the Jeffersonian vision of diversity, openness, and decentralization of control
happen by itself?
The optimist in me thinks we should give telephone and cable companies every
opportunity to get it right. More precisely, we should seek to educate and
enlighten, while developing contingency plans. Here are some principles that
should guide regulators:
- Encourage competition. Competition does more to keep firms honest than
a roomful of regulators.
- Keep government intervention where it is necessary to a minimum. It should
not take the form of heavy-duty regulation of operations, which is self-defeating,
but of the adoption, oversight, and, if necessary, enforcement, of certain
- Ensure free speech and privacy. Constitutional protections of personal
privacy and freedom of expression should be extended to the emerging networks.
As to the networks themselves, they must be built as open systems intended
The Jeffersonian ideal a diversity of users and manufacturers, grassroots democracy
with true communications among the people, and all the dazzling goodies of home
shopping, movies on demand, teleconferencing, and cheap, instant databases is
composed of high bandwidth, distributed two-way switching, and an open architecture.
It's our choice to make. Let's not blow it.
- Easy Access. Everyone should be able to connect.
- Diverse Content. Users should be able to determine the content of the
- Multiple Uses. People should be able to choose the roles they wish to
play as consumers, providers, or both.
- Flexible Architecture. Networks must be built as a series of interoperable
components with well-defined published interfaces which permit maximum third