High speed networking is the sinew and muscle of the New Economy, we are told. The Internet is but the latest in a long history of innovations which have shrunk distance: the cog (the medieval transport vessel which enriched northern Europe by enabling reliable, low cost long distance trade), canals, the railway, the steamship, telegraph, telephone, radio, internal combustion engines, interstates and turnpikes, commercial air travel, television.

The interesting thing is that virtually each innovation was introduced as a monopoly: medieval robber barons erected tolls every few miles along the roads and waterways of Europe; long distance sea routes were protected first by secrecy, then by the granting of trade monopolies; the Railway Barons of America created interlocking empires of rail, oil and finance; telegraph companies resorted to chicanery, boycott and even violence to protect themselves from the disruptive technology of the telephone; telephone companies themselves became jealous monopolies, often folded into the postal services (the PTT model still obtains in much of the world).

For us in Atlantic Canada, the past has several important lessons.

Transport and communications are a single continuum. While this was always true, it has become more poignant now that information rather than goods provides the greatest part of wealth creation and job creation.

The traditional model for management of the transport-communications continuum has been the “Scarce Resource Model”. Particularly prevalent in Canada, where huge distances, sparse population and harsh weather conditions add to the challenges, the scarce resource model assumes huge infrastructural investment must be encouraged by monopoly protection, that scarce resources will be rationed by price and that the public interest will be protected by regulation.

In fact, the model fails as the previously scarce resource becomes commoditised, although the monopolists seek to protect their position by regulation and legislation (limiting access to competing technologies) and capture of the regulating bodies. New, disruptive technologies route around the entrenched incumbent. We saw this process as telephone overtook telegraph and as road bypassed rail. We are seeing the process at work as packet-switched data services route around the circuit-switched voice service-based telephone companies.

Governments have usually sought to protect incumbents against the effects of disruptive new technologies. The reasons are self-evident: monopoly suppliers have built powerful political tools to protect and justify their positions and privileges; they represent large scale, well paid employment and a large tax base; being bureaucracies themselves, the are easy and convenient for government bureaucracies to interface with.

It is essential to the “utility” model that the resource being managed continue to be treated as an expensive scarce resource – even if it has effectively been commoditised. The public interest, however, is best served by the commoditisation of the resource in both management and pricing, not just technical availability. The Island’s air connections provide a powerful case in point. The fixed link is another example where a previously scarce resource has been commoditised (the marginal cost of one more car crossing the bridge is effectively zero) yet is managed on a monopoly-controlled scarce resource basis.

The Atlantic region, being at the periphery of most transport-communications infrastructures, is particularly vulnerable to monopoly abuse. For over a century, successive monopolies have controlled the region’s access to its markets – particularly PEI’s, where even the road infrastructure is interrupted by monopolies governing access to the mainland.

The impact of monopoly (and the scope for abuse) is significantly enhanced when the monopolist is both supplier of the underlying infrastructure and a competitor to supply services on that infrastructure. Around the world, there have been attempts to liberalise monopolies by allowing others to supply services while leaving the incumbent monopolist in control of the underlying infrastructure. The telephone network is perhaps the most prominent, but other examples are to be found in railways, electricity and gas. Inevitably the incumbent is incented to abuse his control (even where mediated by an “independent regulator”) of the infrastructure to protect his position as a supplier of services over that network – the more so as the services (without effective competition) can be sold more profitably than the underlying basic access product. The difference in margins between providing a household phone (price regulated and availability guaranteed) and offering long distance services (especially before vigorously contested competition finally began driving down charges) is self-evident.

The role of a regulator becomes confused when disruptive technologies offer alternate solutions. Always vulnerable to political pressures and capture by the entity being regulated, the regulator is usually anxious to extend his control to the new technologies. Whether the French language police of Québec or the CRTC, regulators look at IP networking as a chaotic and unruly newcomer in obvious need of their guiding hand. To date, the guiding maxim of the Internet – the Web routes around censorship – has largely foiled the efforts of the regulators in terms of content. However, much of the physical infrastructure of the Internet has fallen under their purview because it runs over telephone lines.

The fact that the Internet happened to run over telephone lines is an historical quirk. It was originally designed to provide a failsafe communications infrastructure in the event of an electro-magnetic pulse from a nuclear weapon knocking out the “smart network” of the telephone companies.

There is no logical reason why telephone companies should dominate or own the infrastructure of the Internet. There is probably less in common between the technologies of the telephone network and the Web than there was a century ago between the technologies of the telegraph and the telephone. In fact, for years the telephone companies ignored the Internet as an irrelevance, referring to a scum of data floating on a sea of voice. Two years ago, the volume of data traffic overtook that of voice on the US telephone system.

For years, the primary objective of the telephone companies was to protect their investment in expensive, circuit-switched or “smart” networks. The natural environment of the Internet is low cost, packet-switched or “dumb” networks. There is a natural analogy between the railway system, with its switches and preordained routing of traffic and the highway system where drivers independently select the best route to their destination and drive around bottlenecks.

Now telephone companies have begun to accept the new packet-switched paradigm for their future 3G (Third Generation) wireless networks, but all the indications are that instead of seeking to commodities data traffic on those networks and drive costs down, they want to provide data as a premium service carrying additional tariffs.

The introduction of the fax gives a clear picture of how a relatively unregulated network economy operates. The first readily available fax machines cost on the order of $7,500 and a business might discover that only one or two out of hundreds of customers had one. The unit cost of access to the network was high and the value of the network was low. Two decades later, faxes cost in the low hundreds of dollars or come free as computer software. Hundreds of millions of fax nodes exist. The unit cost of access to the network is tiny; the overall value of the network is incalculable. E-mail drove this lesson in network economics further. The marginal cost of e-mail is effectively zero; the cost does not increase if the number of recipients increases from one to a thousand or even ten thousand.

These numbers fly directly in the face of the scarce resource model. This is the reason that one will hear telecommunications vendors talking of “value-based pricing”. They want to have the right to establish the value of the service to the user and charge accordingly. This is a railway model. By contrast, most users favour a “roads” model of open access. Inevitably that day will come (when governments recognise data networks as an extension of the road system and meet the base costs from general revenue), but it is probably not a viable short-term option, especially given the build costs of providing a new infrastructure. It is interesting that companies such as Cisco and HP are already designing equipment to run on optical networks which can distinguish among kinds of traffic and bill accordingly. It should be noted that this is for networks capable of 40 gigabits a second: a plentiful resource will be tariffed as a scarce resource. Of course, the reason they are doing so is that telecommunications companies are their core customers, and telecommunications companies want to apply the scarce resources model to the optical networks of the future. This suits the Nortels and Ciscos well: telephone companies applying a scarce resources model to their customers will not query the reasons their vendors feel it appropriate to make a 60-70% profit margin on the equipment they sell.

The recent catastrophic falls in the share prices of Cisco and Nortel suggest the market may not share their very comfortable view of broadband rolling out as a technology which has effectively been commoditised, yet is managed as a scarce resource.

A Solution for PEI

Although a “to every door” broadband network for the Island has been under active consideration and discussion for some months, the prospect achieved a clearer focus with the recent announcement by the Minister of Development and Technology. Very interesting initial work by Bill MacPherson of Holland College and others has worked through processes both formal and informal to the extent that debate is now focussed on “how” and “when” rather than “why” or “if”. There appears to be some element of consensus on certain points:

1. Like the telephone system, it should be available to every Islander, whether in Charlottetown or on the top of Elephant Rock;
2. It should be multi-vendor at point of use: in other words, any and all qualifying vendors of services should have equal access to deliver their offerings over the network;
3. There should be strict separation between the supplier of the underlying infrastructure and providers of services on the network. Whoever manages and operates the network should not also be allowed to compete to provide services over it (either directly, or through surrogates). This will eliminate a major conflict of interest that has significantly inhibited IT development in the province.

On some equally important issues, there is not yet consensus:

1. Who should own and manage the service?
2. Should the rollout be incremental, beginning with the public sector and particularly the educational establishment?
3. Should the network be conceived ab initio as permitting full convergence of data/computing, telephony and broadcast?
4. What protections should be built in to protect existing suppliers of services?
5. How can the network best be leveraged to achieve wider objectives; for example, to build on the significant investments already made in Community Access Projects (CAP), Access.Ca and Access.PEI and School.Net? To contribute to broader economic development, particularly of the IT sector?
6. What regulation applies? What is appropriate/necessary?
7. What will be the pricing model?

We have a vision of Prince Edward Island as the first jurisdiction in the world to roll out a fully converged high-speed network with low cost access available to every household, business and organisation. We see a partnership between the public and private sectors that builds on the self-evident strengths of each, while at the same time avoiding the weaknesses and disadvantages of each. We see a partnership between federal and provincial governments, building on our role of prototyping technology solutions for later rollout across Canada (e.g., CAPs, Access.Ca), driving the federally announced objectives of broadband to every community in Canada by 2004 and the delivery of all services to every citizen via the Web by 2004. We see utilising the one-off benefits of the Tomorrow’s Wave investment in the economic development of Atlantic Canada to build a resource our grandchildren will thank us for.

To achieve those objectives requires creative and constructive thinking. Certain principles seem clear to us:

1. Our strength as a community is our ability to work bottom up, for the community and its institutions to mobilize to achieve ambitious objectives. Just as our grandparents built the school systems with adzes and awls, we should build our network from the ground up.
2. There are two basic, competing models of computer networking. These can be summarised as below. While each value pair is actually a continuum, the objectives most to be desired will be achieved by optimizing for the characteristics in the right hand column:
   
    

3. For years, the first priority in providing broadband network services has been the public sector. The outcome has been a highly expensive network (or group of networks) that the taxpayer has paid for, but the vendor owns. Although this has been bragged on as a public network from time to time, it is in fact a private network for the public sector and has been largely underutilized. We understand a resource-driven model which suggests building the first tranche of the proposed new network as an educational and science network, but we believe it is essential to engage the public from the earliest stage possible. There are possible options:
   - concatenate the present three stage rollout to begin at least beta public rollouts
   - extend the proposed Science Research and Education network to include CAP2 sites and provide short-term access for private sector businesses within that framework.
4. The Island’s existing CAP2 sites are a potentially powerful resource for the management of the Central Offices (CO) or nodes of the network. This would build on the community access skills and tools they have developed and the major investment made in them by both the federal and provincial governments. As partners in the operational management of the network, they would build up highly exportable additional skills and the basis for an export service of developing low cost broadband networks in remote and rural areas. (Eighty percent of the world’s population currently lies beyond the Internet; the efforts of the major vendors will continue to focus on high-density affluent areas in the cities and suburbs of the developed world.)
5. We understand the motivation that leads the government to look provisionally to the private sector to build, own and manage the full network, while beginning with the model of public ownership of the proposed Science, Research and Education network to “prime the pump”. We favour the idea of the people of the province having a direct ownership stake in the network. The province has a hundred and twenty-eight year history of being dependent on transport-communications infrastructure owned and managed out of central Canada. That history is not encouraging in terms of providing examples where Islanders received high quality services for reasonable cost. The network is also far too strategic a resource in the development of the provincial economy to be entrusted to the unrelieved mercies of the private sector: again, history is instructive. We favour a model of a public/private partnership in which the public sector has a strategic stake hold. We further believe that this could be used at some future time (say within five years) to strengthen the provinces nascent capital markets by sale to the public with the incentive of the province’s Private Investor Equity Incentive tax rebate. Protection could be built in by limiting any holding to not more than 5% of the outstanding equity.
6. We believe that the network should be conceived ab initio as a fully converged network in which data, telephony and broadcast services are available. Inevitably, this will involve the CRTC, although in principle we favour the conclusions of last year that CRTC regulation of data services is not desirable. However, common carrier status is also important to enshrine the right of all Islanders to access the network. The whole issue of regulation will continue to evolve in the medium term as governments struggle to understand the impact of the new technologies.
7. We should be in early discussions with the Department of Industry, ACOA, the National Research Council and other interested parties to maximise the financial resources available and to build deliberately as a prototype for fulfillment of the federal government’s commitment to roll out broadband to every community and Canada and to deliver government services over that network. Partnership is a key working tool.
8. While fibre optic lies at the core of our thinking, there are other distribution technologies will permit early rollout and infill for particularly remote or difficult locations. Wireless is a key such technology, and several wireless technologies – from fixed terrestrial broadband to 2.5G and 3G cellular telephony – may have a role to play. In most communities internationally, hybrid networks will be the norm and our planning should factor in the issues of cost, appropriateness and universality with a view to creating a readily exportable solution.
9. We feel the “scarce resource” pricing model is inappropriate. Our objective is to create a low cost canvas on which vendors, users and partners can supply services and price according to competitive market conditions. Obviously the pricing of basic infrastructure must allow for amortisation, renewal, reinvestment and a reasonable return on capital employed. Because the public sector currently pays very high prices for network services, and would be a major user of the network, it would be a key “anchor tenant” whose continued usage fees would help provide an important element of financial “plannability” and sustainability and whose utilisation of the network to deliver services would drive users. It own network costs would fall on a like-for-like basis, although its current plans call for greater network utilisation. Other potential swing users can enhance service while reducing cost: for example, there are models where the electric utility helps reduce network costs by supplying access to poles while reducing it own costs by using IP instrumentation to read meters, monitor for system faults, etc. This reinforces the argument in favour of a collaborative approach rather than a single vendor model.

Conclusion

The purpose of this paper is to stimulate debate. Too often, key decisions about important infrastructural issues are made behind closed doors in close discussion with vested interests. The government is to be congratulated on having encouraged debate and discussion, and on having encouraged officials to consult widely in the communities of interest. We have tried to outline some of the key issues as we see them affecting the public, the IT sector and broader development policy in the province.

We have a once-in-a-generation opportunity to secure pre-emptive leadership, to create something which is a model, not just for the non-urban areas of Canada, but for much of the world. Never in the history of the province has it actually controlled its links to the outside world. In the New Economy, where the importance of the network will outweigh that of roads or rail, that control it not merely highly desirable – it is, we suggest – essential. It is also completely within our grasp.

The technology is not the issue. It is now commoditised. It is not a scarce resource to be rationed by price. We should regard it as wallpaper – not the Mona Lisa. The huge value in the network will reside in what use our institutions, businesses and communities make of it. They will create the Mona Lisas.