Presented at MIT Workshop on Internet Economics March 1995

Abstract

Research on Internet Economics attempts to improve our understanding of the Internet as an economic system. The lack of accepted metrics for economic analysis of the Internet is increasingly problematic, as the Internet grows in scale, scope, and significance to the global economy. This introduction brings together interdisciplinary research on the Internet—a network of networks employing common standards to achieve the technical goal of statistical sharing, the economic goal of positive network externalities, and the policy objective of interoperability.

Models and frameworks for Internet technical, economic, and policy analysis are presented in this special issue on Internet Economics of the Journal of Electronic Publishing. Research, engineering, policy, and business issues are addressed in six categories: 1) economics of digital networks, 2) Internet resource allocation and pricing models, 3) responsive pricing, 4) requirements for pricing Internet services, 5) network interconnection and Internet economics, and 6) Internet economics and policy.

This article introduces the reader to these topics. We explain why articles in this special issue are grouped in the categories listed above and identify areas for further research to advance knowledge of Internet Economics.

1. Introduction

The convergence of digital technologies including television, telephony, and computers has stimulated the reach of the innovations of the Internet. Digital video, audio, and interactive multimedia are growing in popularity and increasing the demand for Internet bandwidth. As yet, however, there has been no convergence or consensus on either the economics or the future policy framework of the Internet. While advanced information and communication technologies make the Internet work, economic and policy issues must now be addressed to sustain the growth and expand the scope of the Internet. This has led to the emergence of the field of Internet Economics as a subject of academic and industry research.

This article is an introduction to Internet Economics. Prior work has been done on the technical, economic, and policy development of the Internet [Hart, Reed, Bar, 1992]. This publication is one of the first attempts at bringing research on Internet engineering and economics together. The articles in this special issue include architectural models and analyses of Internet usage as well as alternative pricing policies. We also identify areas for further research. We describe the structure of the remainder of the special issue and explain why the articles are grouped as they appear in this journal.

The issues raised by Internet Economics are not new. Accounting and pricing issues have been an issue for computer networks for some time. [1] In 1974, Len Kleinrock wrote about future issues of computer communications [Kleinrock, 1974]. While he focused on the technical problems of scalability and heterogeneity in networks he also raised the central challenges for Internet Economics, [H]ow does one introduce an equitable charging and accounting scheme in such a mixed network system. In fact, the general question of accounting, privacy, security and resource control and allocation are really unsolved questions which require a sophisticated set of tools. While work on the development of the Internet has been largely technical to date, without new research and a deeper understanding of these economic and policy issues, further growth of the Internet may be stymied.

Two commonly held misconceptions about the Internet involve its economics. The first is that it is paid for by the United States government when, in fact, in 1994 the National Science Foundation paid for less than 10% of the Internet cloud [2]. and far less than that percentage of the cost of Internet infrastructure and software development. For most users, the cost of local area networks, computers, and local telephone lines far exceeds the costs that can be ascribed exclusively to Internet connectivity. The second misconception is that the Internet is free. In this journal, MacKie-Mason and Varian detail the frequently asked questions concerning the economics of the Internet and the economics for the infrastructure [MacKie-Mason and Varian, 1993, 1995], and show that while the marginal cost for Internet traffic may approach zero due to statistical sharing, other costs, such as congestion costs, may be significant.

As an investor and large user of the Internet, the U.S. Government, like other governments, is nonetheless concerned with the self-sustaining economic growth of the Internet. The mission agencies, such as NASA and the Department of Energy, operate their own wide area Internet Protocol network, which interconnects with the Internet. Therefore, the government continues to be conscientious about how it invests in them. One of the main documents that outlines this guidance is the Office of Management and Budget’s (OMB) Circular A-130, Management of Federal Information Resources. This circular outlines how federal agencies should coordinate their efforts so that they may provide Internet services in a cost effective manner.

Economically, the Internet exhibits the feature of positive network externalities. Technically, it benefits from statistical sharing of network resources. And it reaches the policy objective of interoperability. [3] Developing a more refined model of pricing the Internet that preserves the low transaction costs enabled by statistical sharing, positive network externalities, and interoperability is the common goal of the papers presented here. Areas of disagreement are evident, particularly on the subject of the costs and benefits of congestion pricing. Nonetheless, we believe the papers in this volume contain the seeds of a model for the Internet’s evolving technical, economic, and policy architecture.

The challenge for government and industry is not only to develop effective Internet pricing mechanisms, but to do so without losing the benefits of interoperability and positive network externalities currently gained through the Internet’s reliance on statistical sharing. Achieving economic efficiency without inciting users to abandon the Internet’s technical approach of statistical sharing, and hence without abandoning the Internet as such, is the challenge businesses, policy makers, and the public must now face.

2. Structure of this special issue on Internet Economics

Research across disciplines on Internet Economics is difficult. While the communities that study, design, and use the Internet have many ideas on Internet Economics, less thought has been put into a framework combining the merits of the various proposed solutions. It is increasingly apparent that this is an impediment to the further development of the Internet. Better understanding of both the engineering and economics of the Internet is needed.

Each section of the journal conveys the most recent and, at times, contradictory thoughts of leading scholars on Internet Economics. The first section contains reviews of the basic economic qualities of digital networks, including the Internet. The section also introduces and places in historical context the debate over flat rate, usage sensitive, and transaction based pricing. The next three sections address Internet pricing—pricing of packet delivery, user responses to pricing, and pricing content. These sections are somewhat overlapping, but are separated to highlight differences in approaches. The last two sections address economics and policy. The first of these two sections examines business policy and corporate strategy concerns with how firms work with one another in interconnecting their networks. The last section looks at public policy issues of regulation and public goods.

2.1. The Economics of Digital Networks

The first articles in this journal are not limited to Internet economic issues but focus more broadly on the economics of digital networks. The first paper by MacKie- Mason and Varian introduces and answers many questions about the economics of the Internet. This paper is organized as a FAQ—an acronym for frequently asked question—so it can easily be scanned for answers to some of the most fundamental questions underpinning the field of Internet economics. For many technologists and economists, reading the MacKie-Mason and Varian FAQ has served as a first introduction to the field of Internet Economics. The process of writing and distributing FAQs is an integral part of the Internet community. FAQs on many topics are on servers scattered around the Internet. The MacKie-Mason and Varian FAQ on economics and the Internet has been the most influential FAQ about Internet Economics; the feedback the authors get from the FAQ’s readers helps them make new and better revisions of their work. We are pleased to include their latest version in this journal.

The remaining papers in this chapter address broad issues of Internet Economics from three different perspectives: architectural, historical, and empirical. The Gong and Srinagesh chapter provides the architectural focus, through an economic assessment of the effect of layering networks— one of the most powerful technological forces behind achieving interoperability in the Internet. Anania and Solomon looked at the development of B-ISDN services in the 1980’s and discussed why flat-rate pricing for services makes sense for users. The paper is reproduced as published in 1988 with a new introduction. Many of the arguments over pricing on future broadband networks parallel the work found in later sections of the electronic journal on Internet pricing.

Finally, we close this section with Brownlee’s empirical case study of Internet pricing in New Zealand. The structure of the Internet allows for innovation—including pricing models—which may be separate from pricing methods of other Internet networks. This is in fact one of the critical challenges for any attempt to implement a settlements process across the entire Internet. New Zealand is one national network where usage-sensitive pricing has been implemented successfully. The New Zealand experience should be kept in mind when reading future proposals to implement usage- sensitive pricing models, while realizing there are reasons, such as New Zealand’s geographic isolation and the expensive trans-Pacific network link, that may limit the applicability of the New Zealand experience and approach.

2.2. Internet Resource Allocation and Pricing Models

Internet pricing is a growing area of concern to members of the technical, business, academic, and user communities. With new applications, new users, and new connections, the Internet has become an important medium for communication, information dissemination, and electronic commerce. But the economic policies currently undergirding the Internet may be less extensible. The papers included in this section of the special issue present new thinking and research on Internet resource allocation and pricing.

Pricing is still a taboo term for some when put in the same sentence with the Internet. It is considered to be against the traditional Internet culture to adopt a pricing scheme where users need to put a postage stamp on every letter (in this case, e-mail), to pay for its delivery. There also has been a lack of understanding and miscommunication over Internet pricing issues. We hope to contribute to clarity in the debate by defining pricing models as belonging to one of three categories:

1. Flat-rate pricing: Currently, most Internet users pay a fee to connect while they are not billed for each bit sent. For example, a user may pay for a T1 link regardless of how many bits they receive or send. [4]
2. Usage-sensitive pricing: Users pay a portion of their Internet bill for a connection (this price could be zero, but rarely is) and a portion for each bit sent and/or received. The marginal monetary cost of sending or receiving another bit is non-zero during some time period. It is possible, for example, to have usage- sensitive pricing during peak hours and flat-rate during off-peak but we define the overall system as being usage-sensitive pricing.
3. Transaction-based pricing: Like usage-sensitive pricing, the marginal monetary cost of sending and/or receiving another bit is non-zero. However, the prices are determined by the characteristics of the transaction and not by the number of bits.

We do not use the term usage-based pricing in the remainder of this article because it has too many conflicting definitions. For example, the cost of a T1 and a T3 link are different and based upon the predicted use of that link, so the price could be construed to be usage-based. However, others have used the term usage-sensitive pricing synonymously with usage-based pricing.

While most current research on Internet Economics analyzes usage-sensitive pricing models, we recognize that the flat- rate pricing model does have tremendous advantages. Administrative overhead for billing systems could be avoided with a flat-rate pricing system. The percent of the overhead is unknown, but perhaps we can learn from the telephony case where the billing overhead is said to account for approximately 50% of the phone bill. [5] Flat-rate pricing does not, however, provide an economic congestion control mechanism for bandwidth resource allocation. This is especially important with development of IPv6, the next generation Internet Protocol. The delivery of multiple service classes across the Internet provides different qualities of service to different packets. If there are no usage-sensitive fees for the packets receiving better service, all users may prefer better service. This arguably would lead to a greater demand for bandwidth than is available.

This set of articles offers the reader a fresh look at usage-sensitive pricing policies from a number of different perspectives. The papers outline new pricing models, most of them usage sensitive, for Internet bandwidth. Each of the papers uses a different approach—an approach that is usually consistent with the author’s background.

Traditionally, the Internet technical community has not concerned itself with pricing. Technologists preferred to implement technical solutions to allocate bandwidth. The current version of Internet Protocol version 4 (IPv4) offers users a best-effort service and the Transmission Control Protocol (TCP) enables congestion control. The Internet cloud is an inherently statistically shared resource of best-effort service. When the shared resources of the Internet become congested, users experience additional latency or delay and TCP reduces its bit rate when sending data over a congested network. Therefore, it hasn’t been necessary to ration a resource using economics since technology has solved this problem.

The multiple service qualities of IPv6 will require packets to be handled differently in some circumstances. As David Clark’s paper suggests, a token bucket system could be used for these multiple services. By flagging a packet which requires a better quality of service, it is possible for the network service provider to manage traffic differently and bill users for their service on a usage- sensitive basis. Clark suggests that users pay for their expected capacity requirements, rather than their actual use, which moves his proposal partially back toward the flat rate pricing approach that Anania and Solomon championed, based on their historical analysis.

Economists have looked at this same issue from a different perspective and with a different set of tools. Instead of having two qualities of service, as Clark proposes, why not have qualities of service that correspond to price and demand? Since bidding models were first suggested, much work has been done to further the research in this area. Gupta, Stahl, and Whinston analyze the effectiveness of usage-sensitive pricing through the use of modeling. The work of Wang, Sirbu, and Peha is also included to address the issue of pricing for asynchronous transfer mode (ATM) networks. While the Internet protocol is not the same as ATM, ATM technologies are expected to be widely deployed. ATM may be used to overlay transport TCP/IP traffic [Downs, 1994]. This paper is a fresh look at pricing the Internet because it considers the possibility of pricing the Internet at a layer other than the IP layer and it uses a protocol designed for multiple classes of service (unlike IPv4). It may be unfair to characterize these papers as the economist’s point of view since they both include a good understanding of the technology. We do believe, however, that they have taken the notions of economic efficiency through pricing models and have applied these ideas in their papers.

Finally, we should note that there is much more work that needs to be done under the section of pricing and resource allocation. While the Crawford article is a fresh look at the topic and the Kelly paper does a good job of introducing statistical methods for understanding pricing, more needs to be done to resolve the current split between the flat-rate and usage-sensitive camps. While the models do help, implementation and empirical data would do much to inform the debate. There are many implementation issues with any of these models (as the authors discuss in their papers) and it may be that through more trials we can begin to understand how best to implement some of these models. Particularly interesting is how users will react to new pricing policies. Eventually, users, or the market if one prefers, will decide what Internet pricing approach to follow.

2.3. Responsive Pricing

Users are ultimately the ones who feel the effects of congestion and will determine which pricing policies are acceptable by their purchase and use decisions. Even with flat-rate pricing, users who do not wish to experience shorter delays choose to use the network when it is less congested. Therefore, we can think of the Internet and congestion control as having a feedback effect on the user. However, with flat-rate pricing, the users’ only choice is to change their behavior by either sending/receiving or not sending/receiving data. The implementation of usage- sensitive pricing will increase the number of indicators users can give to their network provider and, furthermore, increase the complexity of the feedback process.

Two papers in this journal map the effects of user responses as a function of feedback from network congestion. Danielsen and Weiss propose a mechanism to control how much money users spend on their usage-sensitive traffic. MacKie- Mason, Murphy, and Murphy take a broader view of feedback response and try to explain how feedback to the user changes the aggregate user behavior.

Both papers are leading the way to further studies concerning user response and feedback in Internet pricing models. Perhaps the clearest feedback will come when some of the new usage-sensitive pricing policies are implemented. Then questions regarding the Danielsen and Weiss mechanism can be answered and the feedback loop modeled by MacKie- Mason, Murphy, and Murphy can be tested.

2.4. Requirements for Pricing Internet Services

Pricing questions extend beyond the market for Internet bandwidth and into discussions of the market for information. While research on electronic commerce and information security has developed relatively separately from the work of the previous sections on pricing bandwidth, we believe there will be more similarities than differences in the future. For example, billing and settlement processes developed for information goods likely will be the same infrastructure used to bill for Internet bandwidth. While the work of integrating billing and resource allocation may occur at some future time, the framework for electronic commerce and information security is being developed today.

Electronic commerce and information security are tightly coupled and should be examined together. McKnight et al. describe why this is the case. The authors provide an overview of how electronic commerce and information security has developed recently. The main thrust of this paper is that security becomes an enabler for electronic markets. The Internet is by definition an open, interoperable system designed around systems, so it becomes difficult to communicate in a closed and secure fashion. There is no one entity on the Internet that everyone can trust to provide security. Therefore, how can markets that may require secure and private communication and authentication proliferate? Developments in encryption, such as public key encryption, have helped give business and government hope that security problems will be solved. But what about other underlying frameworks from which to build electronic commerce? The wiliness of hackers and the temptation for fraud, not to mention concerns with terrorism, suggest that information security on the Internet must provide for technological heterogeneity and a diversity of user requirements. Current U.S. federal policy and global Internet business practices still fall far short of this goal.

The two other papers in this section present examples of the mechanisms likely to be required for electronic commerce and information security. The Lai, Medvinsky, and Neuman paper presents examples of licensing and reimbursements in non-electronic markets and creates analogous models for electronic markets. The Neuman and Medvinsky paper on NetCash and NetCheque discusses examples of other components of an electronic commerce system on the Internet using different mechanisms for payment. Both papers address the electronic commerce problem from a security point of view by discussing methods for establishing trust between the supplier and provider of a good.

The true economic benefits of the Internet will likely appear only when electronic commerce is as pervasive on the Internet as in other marketplaces. The foundations that are being laid today, as seen from the papers in this section, should bring us closer to realizing a future of greater network externalities and lower transaction costs. There is a need for further research on characterizing network externalities and transaction cost benefits of Internet services in light of the developments of technologies for electronic commerce and information security.

This is the last section that deals with issues at the finest level of granularity—packets of data and individual users. The next two sections deal with issues of Internet Economics for aggregations of users and at the firm level.

2.5. Network Interconnection and Internet Economics

This section explores the economics of interconnection for firms providing Internet services. Internet service providers (ISPs) are in the rapidly growing business of providing network connectivity to users. Until now, the relative homogeneity of users and the Internet culture of cooperation has given firms incentives to interconnect. However, this is changing rapidly as ISPs are providing competitive services and are appealing to a variety of niche markets.

The papers in this section analyze the economic factors encouraging firms to interconnect. Models for future interconnection agreements are also presented. The Srinagesh paper characterizes Internet interconnection agreements as they exist today and suggests how they may be changing. The data in this paper is insightful, and represents one of the few empirical studies on Internet Economics yet done. The Bailey paper uses incomplete contract theory to suggest why some types of interconnection agreements are more prone to opportunism than others. Both papers offer new methods for Internet network connection and weigh the pros and cons of adopting specific models.

Perhaps the greatest economic concern for network and service providers with interconnection, and one that needs greater exploration, is resale. Economic payments for interconnection can either encourage or discourage resale. [6] For example, with a flat-fee interconnection agreement, there is the possibility of two firms aggregating their traffic before the interconnection point and splitting the fee of the interconnection. Therefore, interconnection points with flat-fee settlements may provide improper incentives for use and may lead to inequitable cost recovery by the interconnected firms. It may be possible, however, to design an economic model that employs usage-sensitive pricing to discourage resale. This may encourage aggregated user behavior to change as well.

When firms do not cooperate and social welfare is threatened, governments may intervene. Regulation, or even criminal prosecution for violation of anti-trust and restraint of trade laws, may occur if a service provider, for example, excluded some firms that wanted service. The role of the government in Internet economics is addressed in the last section.

2.6. Internet Economics and Policy

Public policy affecting the Internet has focused on three distinct areas: regulation of the communications infrastructure, funding for Internet networks, and public investment in science and technology for digital computing and communications systems. One example is the high leased line rates imposed by government ministries, which has been identified as a cause of the slow growth of the Internet in countries such as Japan in the early 1990s. The future development of Internet economics will also be affected by government actions. The two papers in this section discuss two of these roles of government. [7]

Regulation of the Internet may come in many forms ranging from common carriage rules regulating the sale of circuits to content regulation. Sarkar’s paper focuses on market failures of the Internet and makes a case for regulation. Arguably, the Internet is already regulated since many of the telecommunications circuits that make up the Internet are regulated, for example in the United States, at rates set at the state or federal level. Regulation may occur, for example, if Internet service providers establish unfair pricing practices or interconnection agreements.

The paper by Hallgren and McAdams explores the economic benefits of supporting public goods for further development of the Internet. They argue that the market may not provide the right incentives for private development of tools to help the Internet grow and argue that this is the responsibility of governments. Certainly, the U.S. federal government has taken lead roles previously with the development of the ARPANET and the NSFNET. The authors make a strong case for federal leadership for advanced technology, such as that supported by the U.S. High Performance and Computing Initiative. The previous section on electronic commerce and information security however, shows that government regulation may inhibit innovation, and limit the growth of firms. The Internet by definition extends beyond borders, of both nations and institutions. How this remarkable global resource may continue to grow and extend its services is certainly a matter for public concern, if largely to be resolved through private actions of individual users as well as firms.

3. Conclusion

This collection of articles on Internet Economics introduces models of Internet pricing intended to preserve the low transaction costs enabled by statistical sharing, positive network externalities, and interoperability. There is as yet little consensus on the merits of the proposed Internet pricing models. We have included the Internet Economics Workshop notes as an Appendix to this journal, to provide the reader with a less-filtered perspective on the debate under way among Internet Economics researchers from different disciplines. We hope this volume motivates the authors or others to take the next step beyond building models to implementing, evaluating, and comparing proposed approaches to Internet Economics. The effects of user behavior on the performance of alternative models of the economics of the Internet may then be assessed.

Each of the sections that follow in this journal represents much of the best work yet done in areas of research and development which are still in their infancy. While this publication is a step towards advancing the field of Internet Economics, much more is needed for a sufficient understanding to determine business strategies or public policies.

Principles of Internet Economics will be developed as economists, engineers, and others collaborate on further research. We believe there are at least five areas of opportunity where new solutions to Internet economic problems may be explored:

1. New empirical work is needed especially with respect to user behavior. Combinations of flat rate bandwidth pricing with transaction-based pricing is a promising area for further research and commercial interest.
2. Congestion control protocols and user feedback should be tested in concert with one another. This and other integrated approaches to Internet Economics problems should be explored further.
3. Pricing as a congestion control method or as part of an interconnection settlements process should be tested in concert with enhancements to the TCP/IP protocol suite.
4. Increased data collection across Internet service providers will provide an understanding about the growth of the Internet and provide a rich data set to develop economic models and increase understanding of the subtle interaction between Internet engineering, economics, and cultural practices.
5. Further research on the development of technologies and policies for information security for electronic commerce and privacy protection is needed. Preliminary evidence suggests that further technological advance of the Internet will be hampered by lack of adequate attention to these issues.

The Internet has become a powerful innovation engine for the emerging Global Information Infrastructure. To encourage further research, advance the public policy debate, facilitate business decision making, and support technological advance, we offer this collection of some of the best current thinking on Internet Economics. We hope this special issue contributes to a better understanding of Internet Economics and leads to stronger economic performance while safeguarding the Internet culture.

Acknowledgments

Preparation of this special issue of the Journal of Electronic Publishing was made possible by a grant from the Sloan Foundation. The findings discussed in this paper are partially a result of the authors’ work on an Internet Economics Workshop sponsored by the National Science Foundation grant #NCR-9509244. Research support on this topic has been provided by the Advanced Research Projects Agency, contract #N00174-93-C-0036. We would also like to thank the Federal Networking Council and its advisory committee for their interest in and criticism of our work, which was presented at meetings of the FNCAC on April 4 and October 11, 1995. The MBone video distribution of the workshop was made possible by Silicon Graphics, Inc. and the Digital Equipment Corporation.

The Internet Economics Workshop was held in Cambridge, MA at MIT on March 9 & 10, 1995. [8] We wish to thank the participants in that workshop and particularly the organizing committee of David Clark, Deborah Estrin, Jeffrey MacKie- Mason, and Hal Varian, without whose assistance and creative thinking this work would not be possible.

The views expressed in this paper are those of the authors, and not the institutions with which they are affiliated.

Author Information

Lee W. McKnight (mcknight@rpcp.mit.edu) is Associate Director of the MIT Research Program on Communications Policy, Lecturer in the MIT Technology and Policy Program, and Principal Research Associate at the MIT Center for Technology, Policy, and Industrial Development.

Joseph P. Bailey (bailey@rpcp.mit.edu) is a doctoral candidate in the Technology, Management and Policy Program at MIT and a research assistant with the MIT Research Program on Communications Policy.

Dedication

This special issue on Internet Economics is dedicated to Bob Kahn, Vint Cerf, and the few hundred other pioneering fathers and mothers of the Internet. Thanks!

References

Bailey, J. S. Gillett, D. Gingold, B. Leida, D. Melcher, J. Reagle, J. Roh, R. Rothstein, and G. Seale, Internet Economics Workshop Notes, Research Program on Communications Policy, MIT, March, 30, 1995.

Bohn, R. H.-W. Braun, K. Claffy, and S. Wolff, Mitigating the Coming Internet Crunch: Multiple Service Levels via Precedence, Applied Network Research Technical Report, San Diego Supercomputer Center, 1994.

Downs, S. Asynchronous Transfer Mode and Public Broadband Networks, Telecommunications Policy, v 18, no 2, March 1994. [ doi: 10.1016/0308-5961(94)90046-9 ]

Hart, J.A. R.R. Reed and F. Bar, The Building of the Internet: Implications for the Future of Broadband Networks, Telecommunications Policy, November 1992. pp. 666-689. [ doi: 10.1016/0308-5961(92)90061-S ]

Kleinrock, L. Research Areas in Computer Communication, Computer Communication Review, v. 4, no. 3, July 1974.

MacKie-Mason, J. and H. Varian, Pricing the Internet, in B. Kahin and J. Keller, eds. Public Access to the Internet. MIT Press, 1995.

MacKie-Mason, J. and H. Varian, Some Economics of the Internet, University of Michigan, 1993.

McKnight, L. and J. Bailey, eds. Special Issue on Internet Economics, Journal of Electronic Publishing, Fall 1995.

OMB, Management of Federal Information Resources, Executive Office of the President, Office of Management and Budget, Circular A-130. Washington, DC, July 15, 1994.

Notes

1. Of course accounting and pricing issues have long been of concern for telecommunications networks and services.(Ramsey, F.P. 1927. A Contribution to the Theory of Taxation, The Economic Journal. pp. 47-61.) For much longer, accounting and pricing has been a critical function of businesses in market economies. (Stickney, C. Weil, and Davidson. 1991. Financial Accounting: An Introduction to Concepts, Methods, and Uses. San Diego: Harcourt Brace Jovanovich.)

2. The term cloud refers to the network of networks (owned and operated by multiple partners) which provides data transport through use of TCP/IP.

3. Lee McKnight first proposed that the Internet could best be understood as operating on the principles of its positive network externalities, statistical sharing, and interoperability at the Internet Economics Workshop.

4. A T1 link is a circuit between two points at 1.544 megabits per second. Joseph Bailey first introduced these definitions in his presentation at the Internet Economics Workshop.

5. This purported statistic was discussed during the Internet Economics Workshop. See the appendix to this journal (the workshop notes) for the documented discussion.

6. Resale is another issue which has long been of concern to telecommunications firms and governments. For example, until recently, resale of telecommunications circuits was prohibited by international treaty. Resale is now prevalent, and has not had the disastrous effects predicted by early simplistic assessments. In fact, resale can be argued to be a significant positive factor in promoting deregulation and competition and encouraging efficiency. See Bridger M. Mitchell and Ingo Vogelsang Telecommunications Pricing; Theory and Practice, pp. 141-142.

7. Other issues, such as the nature and importance of Internet culture, were discussed at the Workshop but are not addressed in most of the papers collected here. See the Internet Economics Workshop Notes in the Appendix of this volume for information on how the subject of the Internet culture kept popping up during discussions of Internet Economics. Privacy, child protection, pornography and censorship issues which are receiving significant public and government attention — but have not yet been the subject of replicative empirical research — are also beyond the scope of this paper and this special journal issue on Internet Economics. A recent treatment of some of these public policy issues is B. Kahin and J. Keller, eds. Public Access to the Internet. MIT Press, 1995.

rpcp.mit.edu/Workshops/cfp.html]) for workshop notes and other workshop information.