Internet Marketing · eMarketing · Internet Advertising · Online Branding Search Engine Optimization & Marketing · Naming · Privacy Policy · Home · Contact Us Our Services

	ADVERTISING BY ICONOCAST Optimism, Happiness, Self Esteem, Creativity, Competency, Achieve Intrinsic Motivation; visit www.motivationbook.com Iconocast is about learning and teaching without borders; we offer eMarketing, Internet Advertising, Internet Marketing, Search Engine Optimization, Search Engine Marketing, Online Branding, and eMarketing News Services.

Please Do Not Read The Text Below.

Download the original file at

Intermediation and Electronic Markets: Aggregation and Pricing in Internet Commerce
Joseph P. Bailey
M.S., Stanford University (1993)
B.S., Carnegie Mellon University (1992)

We copied the scrambled text from pdf file (that's how it gets copied, sorry we had no control) so the keywords would be searchable.

 

Intermediation and Electronic Markets:
Aggregation and Pricing in Internet Commerce
by
Joseph P. Bailey
M.S., Stanford University (1993)
B.S., Carnegie Mellon University (1992)
Submitted to the Department of Electrical Engineering and Computer Science
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy in Technology, Management and Policy
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
May 1998
© Joseph P. Bailey, MCMXCVIII. All rights reserved.
The author hereby grants to MIT permission to reproduce and distribute publicly paper and electronic
copies of this thesis document in whole or in part, and to grant others the right to do so.
Signature of Author_____________________________________________________________________________
Technology, Management and Policy Program
Department of Electrical Engineering and Computer Science
May 20, 1998
Certified by___________________________________________________________________________________
Lee W. McKnight
Lecturer, Technology and Policy Program
Certified by___________________________________________________________________________________
Erik Brynjolfsson
Associate Professor, MIT Sloan School of Management
Certified by___________________________________________________________________________________
David D. Clark
Senior Research Scientist, Laboratory for Computer Science
Accepted by___________________________________________________________________________________
Richard de Neufville
Professor and Chairman, Technology, Management and Policy Program
Accepted by___________________________________________________________________________________
Arthur C. Smith
Chairman, Department Committee on Graduate Studies
Intermediation and Electronic Markets:
Aggregation and Pricing in Internet Commerce
by
Joseph P. Bailey
Submitted to the Department of Electrical Engineering and Computer Science
on May 20, 1998, in partial fulfillment of the requirements for the degree of
Doctor of Philosophy in Technology, Management and Policy
Abstract
The Internet continues to grow as a medium to support commerce. Economic analysis of Internet commerce is still
in a nascent stage while Internet technology and use has rapidly advanced. The result is an Internet marketplace
brimming with entrepreneurs and major corporations experimenting with business strategies and technology
advances even though the economics of Internet commerce is not well understood. This thesis responds to this need
by exploring how the Internet reduces the market friction common in physical commerce.
The intermediaries who help reduce market friction in physical markets may be eliminated, when suppliers and
consumers increasingly rely on the Internet as a transaction medium. An intermediary in any market may reduce
transaction costs by performing four roles: aggregation, pricing, search, and trust. The intermediary roles of
aggregation and pricing may provide little or no value as the Internet becomes the medium for commerce because
the technology, not the intermediary, reduces transaction costs. This thesis examines the possible elimination of
aggregation and pricing intermediaries in Internet commerce. It does so by extending the theory of the economics of
intermediation and electronic markets; developing a methodology for the analysis of Internet price competition;
analyzing exploratory empirical data of the book, compact disc, and software markets to test a subset of this theory;
and exploring the public policy implications of Internet price discrimination. The approach is interdisciplinary
because this thesis integrates the technology, policy, and economics that underpin the role of aggregation and
pricing intermediaries in Internet commerce.
The thesis shows that Internet commerce may not reduce market friction because prices are higher when consumers
buy homogeneous products on the Internet, and price dispersion for homogenous products among Internet retailers
is greater than the price dispersion among physical retailers. Internet retailers—even those selling homogenous
goods—can develop pricing strategies to differentiate themselves from their competitors and to price discriminate.
The ability for the Internet to become a medium for price discrimination is an area that requires the attention of
public policy makers. While self-regulation of Internet price discrimination may be the most appropriate policy,
monitoring by the United States government through the Federal Trade Commission and international organizations
such as the World Trade Organization helps establish a trusted transaction environment for future Internet commerce
growth.
Thesis Committee:
Lee W. McKnight
Lecturer, Technology and Policy Program
Erik Brynjolfsson
Associate Professor, MIT Sloan School of Management
David D. Clark
Senior Research Scientist, Laboratory for Computer Science
To Wendy
Acknowledgments
This thesis is made possible with the support of many advisors, scholars, friends, and family. My doctoral
committee of Erik Brynjolfsson, David Clark, and Lee McKnight impacted the thesis research direction,
methodology, and analysis most directly. I thank them for their continued support throughout my MIT career and
for their ability to be a source of guidance. While I must share the credit for the strengths of this thesis, all
remaining errors and shortcomings are my responsibility only.
Beyond the immediate academic support of my committee, I have had the pleasure to work in an environment that
encourages hard work, original thinking, and clarity in communication. While it is impossible to name all scholars
in this category, I wish to particularly acknowledge my colleagues, Judy Cardell, Sharon Gillett, Petros Kavassalis,
Tom Lee, Suzanne Neil, Russell Rothstein, Marshall van Alstyne, and Mort Webster. Many of these people are
affiliated with the MIT Research Program on Communications Policy that has been a great environment for research
and networking. I thank all of the RPCP administrative assistants, Gill Cable-Murphy, Julia Malik, and Elizabeth
Yoon, and graduate students I have had the pleasure of working with. Furthermore, my research has benefited from
the many conferences and workshops including the 1995 Internet Economics Workshop at MIT, the Information
Infrastructure and Policy conferences at Harvard, and the Telecommunications Policy Research Conference. My
previous academic work at Carnegie Mellon and Stanford well prepared me for my future studies. Carnegie Mellon
professors Indira Nair and Marvin Sirbu have been particular sources of support. As I prepare for my future
academic work at the University of Maryland’s Robert H. Smith School of Business, I thank my colleagues there for
helping me transition from my life as a student to my life as a professor.
Some chapters in this thesis have benefited from the contributions of others. Tom Lee gave me detailed comments
on Chapters One and Two. The students of the MIT Sloan class 15.967, Electronic Commerce/Marketing and the
Internet, and my collaboration with Yannis Bakos helped with Chapter Three. Chapters Four and Five benefited
from my collaboration with Erik Brynjolfsson and Mike Smith (Bailey, Brynjolfsson, and Smith 1998) and was also
aided by my work as a consultant to the Organisation for Economic Co-Operation and Development (OECD 1998).
Robert Angulo, Jeff Chow, Husham Sharifi, Marc Shuster, and Jim Wahl assisted me with the data collection in
Chapter Five. Bill Drake and Bill Lehr helped with the substance in Chapter Six. The writing style in all chapters
benefited from the copy editing of Elizabeth Yoon and John Nicholson.
The research environment and educational process of earning a Ph.D. was made possible thanks to the sources of
funding. My research has benefited from numerous sources over time including DARPA (contract #N00174-93-C-
0036), NSF (grant #NCR-9509244), OECD, MIT Center for Coordination Science, and BT. The fellowship from
NASA (grant #NGT-51407) for the last three years of my work was particularly helpful.
My family and friends are sources of love and joy in my life. Because Wendy, my wife, and I come from large
families, I cannot name all family members I wish to thank. However, I have certainly felt their influence. My
parents, Kenneth and Margaret, are my true role models in life and I hope to follow in their example of being both
good at life and good at work. My siblings, Patricia and Michael, have also benefited from my parent’s love and I
know we will always be close friends as well as close siblings. My grandparents, Katherine and Joseph Korb, and
Ruth and Stanley Bailey, have always been sources of inspiration for me. Wendy’s family has become my family so
Ginny Miller, and Marnie, John, Peter, and Anne Nicholson deserve particular thanks. My friends in school and in
life are very important to me. Tom Lee, Dragos Maciuca, Jeff Shelton, and Shane Siefer, the groomsmen at my
wedding, are true friends and I ask them to accept my thanks on behalf of all my friends.
As I was searching for a Ph.D. program five years ago, I was looking for an advisor to guide me through the rough
waters of the doctoral process. I was very lucky to find such and advisor in Lee McKnight. From the very onset he
was very enthusiastic about my research. I am very happy to have found such a great advisor and friend in Lee. He
and his wife Elaine have been great champions for me in my academic career and I thank them very much.
To my partner in life Wendy, I thank you for your unfailing love and support. You are my best friend and greatest
source of inspiration.
Table Of Contents
1. INTRODUCTION.............................................................................................................................................. 9
2. CONTEXT AND LITERATURE REVIEW .................................................................................................. 13
2.1. TRANSACTION COSTS AND INTERMEDIATION .............................................................................................. 13
2.1.1. Transaction Costs in a Disintermediated Market................................................................................ 14
2.1.2. Transaction Costs in an Intermediated Market................................................................................... 16
2.2. MENU COSTS AND RATE OF CHANGE........................................................................................................... 19
2.3. ELECTRONIC MARKETS AND INTERNET COMMERCE.................................................................................... 20
2.3.1. The Nascent Stage of Internet Commerce ........................................................................................... 25
2.3.2. Transaction Costs and the Internet ..................................................................................................... 27
2.3.3. Menu Costs and the Internet ............................................................................................................... 30
3. INTERMEDIARY ROLES AND THE INTERNET ..................................................................................... 33
3.1. INTERMEDIARY ROLES................................................................................................................................ 33
3.1.1. Aggregation........................................................................................................................................ 36
3.1.2. Pricing................................................................................................................................................ 37
3.1.3. Search ................................................................................................................................................ 38
3.1.4. Trust ................................................................................................................................................... 39
3.2. INTERMEDIARIES IN MARKETS AND HIERARCHIES....................................................................................... 40
3.3. INTERMEDIATION AND THE INTERNET.......................................................................................................... 50
4. INTERNET PRICE COMPETITION METHODOLOGY .......................................................................... 57
4.1. HYPOTHESES DEVELOPMENT...................................................................................................................... 58
4.1.1. Bertrand Competition......................................................................................................................... 58
4.1.2. Price Changes: Signaling and Response............................................................................................ 68
4.2. ANALYZING THE DATA............................................................................................................................... 69
4.2.1. Statistical Data Analysis with Normal Distribution ............................................................................ 69
4.2.2. Nonparametric Analysis...................................................................................................................... 71
4.2.3. Analysis With Endogenous Consumer Heterogeneity ......................................................................... 73
4.3. DATA GATHERING ...................................................................................................................................... 75
5. AN EXPLORATORY ANALYSIS OF INTERNET COMMERCE............................................................ 79
5.1. DESCRIPTION OF THE EXPLORATORY DATA SET.......................................................................................... 81
5.2. EMPIRICAL ANALYSIS OF EXPLORATORY DATA .......................................................................................... 83
5.2.1. Higher Prices on the Internet.............................................................................................................. 83
5.2.2. Less Price Dispersion on the Internet ................................................................................................. 88
5.2.3. Product and Market Characteristics................................................................................................... 89
5.2.4. Frequency of Price Changes ............................................................................................................... 93
5.3. AMAZON.COM, BARNES & NOBLE AND COMPETITION IN THE INTERNET BOOK MARKET ........................... 95
5.4. ANALYSIS ................................................................................................................................................. 101
5.4.1. High Search Costs............................................................................................................................ 102
5.4.2. Lack of Trust .................................................................................................................................... 104
5.4.3. Immaturity of Internet Commerce ..................................................................................................... 104
5.4.4. Price Discrimination........................................................................................................................ 105
5.5. LIMITATIONS OF EXPLORATORY ANALYSIS ............................................................................................... 109
6. INTERNET PRICE DISCRIMINATION .................................................................................................... 111
6.1. THE THREAT OF INTERNET PRICE DISCRIMINATION................................................................................... 112
6.1.1. Price Discrimination as a Profit-Maximizing Strategy.....................................................................113
6.1.2. Implementing Internet Price Discrimination..................................................................................... 115
6.2. CONSUMER ACTIONS IN AN UNREGULATED ENVIRONMENT ...................................................................... 118
6.3. PUBLIC POLICY, THE INTERNET, AND PRICE DISCRIMINATION .................................................................. 120
6.3.1. Self-Regulation of Internet Price Discrimination.............................................................................. 121
6.3.2. U.S. Federal Policy ........................................................................................................................... 122
6.3.3. International Efforts to Monitor Internet Price Discrimination........................................................ 125
7. CONCLUSION............................................................................................................................................... 127
7.1. THESIS FINDINGS ...................................................................................................................................... 127
7.2. INTERNET STRATEGY IN A DYNAMIC MARKETPLACE ................................................................................ 129
7.3. FURTHER RESEARCH.............................................................................................................................. ... 130
APPENDIX A. FEBRUARY/MARCH 1997 DATA ANALYSIS ...................................................................... 133
APPENDIX B. MAY 1997 - JANUARY 1998 DATA ANALYSIS .................................................................... 147
APPENDIX C. RETAILER ANALYSIS ............................................................................................................. 151
REFERENCES....................................................................................................................................................... 171
List of Tables
TABLE 3.1: IDENTIFIED ROLES OF INTERMEDIATION .................................................................................................. 35
TABLE 3.2: TRANSACTION COST ESTIMATES FOR MARKET STRUCTURES .................................................................. 47
TABLE 3.3: EXAMPLES OF ELECTRONIC MARKET ORGANIZATIONS AND CONSUMER SEARCH................................... 52
TABLE 4.1: FORMULAS FOR STATISTICAL ANALYSIS OF INTERNET PRICE COMPETITION ........................................... 70
TABLE 4.2: HETEROGENEOUS CONSUMER DEMAND................................................................................................... 74
TABLE 5.1: INTERNET RETAILERS’ RELATIVE PRICE PREMIUM, FEBRUARY/MARCH 1997 ........................................ 84
TABLE 5.2: INSTANCES WHERE MINIMUM PRICE IS FOUND, FEBRUARY/MARCH 1997............................................... 84
TABLE 5.3: MANN-WHITNEY TEST OF MEAN PRICES, FEBRUARY/MARCH 1997 ....................................................... 85
TABLE 5.4: HETEROGENEOUS CONSUMERS’ INTERNET BOOK PURCHASE SURCHARGE ............................................. 86
TABLE 5.5: STANDARD DEVIATION OF PRICE DISCOUNTS, FEBRUARY/MARCH 1997 ................................................. 89
TABLE 5.6: MARKET DIFFERENCES ............................................................................................................................ 89
TABLE 5.7: SUMMARY OF RETAILER PRICE DISCOUNTS ............................................................................................. 92
TABLE 5.8: PRICE CHANGES OF PHYSICAL AND INTERNET RETAILERS, FEBRUARY/MARCH 1997 ............................. 94
TABLE 5.9: CHANGING PRICES IN INTERNET COMMERCE ........................................................................................... 94
TABLE 5.10: 1997 COMPETITION BETWEEN AMAZON.COM AND BARNES & NOBLE................................................... 97
List of Figures
FIGURE 2.1: TRANSACTION COSTS WITH DISINTERMEDIATION................................................................................... 15
FIGURE 2.2: TRANSACTION COSTS WITH INTERMEDIATION......................................................................................... 17
FIGURE 2.3: INTERMEDIATED AND DISINTERMEDIATED VALUE CHAINS .................................................................... 18
FIGURE 2.4: PRODUCT, PROCESS, AND PLAYER “DEGREES” OF ELECTRONIC COMMERCE......................................... 24
FIGURE 2.5: DEFINING INTERNET COMMERCE AND ELECTRONIC MARKETS............................................................... 24
FIGURE 2.6: CORRELATION BETWEEN BUSINESS LOCATION AND SALES .................................................................... 28
FIGURE 3.1: CONSUMER SEARCH FRAMEWORK.......................................................................................................... 42
FIGURE 3.2: DISINTERMEDIATED TRANSACTIONS........................................................................................................ 43
FIGURE 3.3: TRANSACTIONS IN A HIERARCHY............................................................................................................. 44
FIGURE 3.4: TRANSACTIONS IN A MARKET.................................................................................................................. 45
FIGURE 3.5: COMPARISON OF MARKET STRUCTURE VERSUS TRANSACTION COSTS................................................... 47
FIGURE 4.1: DATA GATHERING FLOW CHART............................................................................................................ 77
FIGURE 5.1: MINIMUM PRICES IN THE BOOK MARKET ............................................................................................... 87
FIGURE 5.2: MINIMUM PRICES IN THE COMPACT DISC MARKET ................................................................................ 87
FIGURE 5.3: MINIMUM PRICES IN THE SOFTWARE MARKET ....................................................................................... 88
FIGURE 5.4. PRICE DIFFERENCES AT AMAZON.COM AND BARNES & NOBLE.............................................................. 98
FIGURE 5.5: PRICE CHANGES IN THE INTERNET BOOK MARKET, FEBRUARY/MARCH 1997........................................ 99
FIGURE 5.6: PRICE CHANGES IN THE INTERNET BOOK MARKET, MAY 1997 - JANUARY 1998 .................................... 99
FIGURE 6.1: PRODUCER AND CONSUMER SURPLUS ................................................................................................... 114
FIGURE 6.2: MECHANISMS FOR CONSUMER PROFILING............................................................................................ 116
1. INTRODUCTION
Neoclassical economic analysis often assumes a “frictionless” economy, that is, the ability for all consumers to have
perfect information and zero transaction costs. In such a world, consumers could make perfect economic decisions
because uncertainty is reduced as much as possible–the uncertainty of knowing exactly what the good is, whom they
are transacting with, and if this is the best possible price. In this world, economic theory would describe economic
reality.
However, the world is far more complicated and much more uncertain than the world of the “frictionless” economy.
In the physical world, consumers are limited in their knowledge of what they want, where to buy, and who to buy
from. The difference between the theory explaining the economy and the economy in practice indicates limitations
in neoclassical economics to explain the subtleties of the market. These differences are supported by consumers’
observations in everyday life. For example, the transaction cost of finding and purchasing a given item at a lower
price may be high enough for a consumer to shop instead at a more convenient store, theater, or gasoline station.
Consumers may avoid a restaurant they have never been to because they are uncertain of the quality of the food.
This consumer may choose an inferior-quality restaurant that they frequently patronize because they have imperfect
information. The promise of a frictionless economy is a utopia that seems very far from the grasp of the consumer.
Imperfect information may lead to higher prices. Consumers with misleading or incomplete information often make
decisions that do not maximize their utility from a transaction. Adam Smith observed how information could be
withheld to maintain high prices when he penned the Wealth of Nations in 1776:
“When by an increase in the effectual demand, the market price of some particular commodity happens to
rise a good deal above the natural price, those who employ their stocks in supplying that market are
generally careful to conceal this change. If it was commonly known, their great profit would tempt so
many new rivals to employ their stocks in the same way, that, the effectual demand being fully supplied,
the market price would soon be reduced to the natural prices, and perhaps for some time even below it.”
(Smith 1776, p. 67) [emphasis added]
10 CHAPTER ONE
The use of information to affect price is exacerbated by the fact that today’s global economy is increasingly
becoming more dependent on information. Machlup (1962) and Porat (1977) describe the U.S. economy as an
“information economy” because most American workers gather, process, or create information.1
Information technology can help consumers find information more easily thereby reducing the problems of
imperfect information that maintains higher prices. The information can also be delivered in a timely manner
because of technology. One of the largest contributors and greatest beneficiaries of the information economy is Bill
Gates, Chairman of Microsoft. In his book The Road Ahead, Bill Gates sees the future marketplace living up to the
expectations of Adam Smith:
“Capitalism, demonstrably the greatest of the constructed economic systems, has in the past decade clearly
proved its advantages over the alternative systems. The information highway will magnify those
advantages. It will allow those who produce goods to see, a lot more efficiently than ever before, what
buyers want, and will allow potential consumers to buy those goods more efficiently. Adam Smith would be
pleased. More important, consumers everywhere will enjoy the benefits.” (Gates 1995, p. 183) [emphasis
added]
One particular information technology tool, the Internet, has the potential to shape markets more than any of its
predecessors have. Only recently has the Internet become a magnet for entrepreneurs and Fortune 500 companies
alike who are looking for opportunities to conduct commerce using this interoperable communications
infrastructure. The principles underpinning the Internet are fundamentally different than those on which other
global networks such as the public switched telephone network are based. Specifically, the interoperability of the
Internet allows for communication among users across heterogeneous systems without information degradation.
The result of interoperability is a network that can incorporate change and promote competition. The Internet is still
a dynamic technology but the economics of the Internet will shape future markets.
The Internet may lower the static market friction costs of transacting. As information becomes more widely
available to consumers and suppliers alike, problems of imperfect information are reduced. Consumers may now be
able to make decisions that maximize their utility because information is inexpensive to find and process. These
benefits result in lower transaction costs in a market exchange. Furthermore, suppliers may find it more difficult to
conceal information from their competitors so that price competition is more likely.
The Internet also increases the dynamic nature of the marketplace. Unlike physical markets where change occurs
slowly because of printing and distributing information delays, change occurs very rapidly on the Internet.
Information can be globally distributed electronically in seconds. For markets, the Internet is a tool for almost
instantaneous consumer feedback. Furthermore, this feedback can be processed and acted upon quickly and at a low
1 More recently, Tapscott (1996) uses the term “ digital economy” and Vairan (1996) uses the more traditional term “ information economy” to
describe how the Internet increases the importance of the information sector to the U.S. economy.
INTRODUCTION 11
cost relative to physical markets. For example, prices can be changed dynamically to meet demand because the cost
of changing a price—the menu cost—may be lower on the Internet than in physical markets. Dynamic pricing may
lead to more price competition among Internet retailers because they can respond to their competitor’s actions more
quickly. It may also lead to a strategy of Internet price discrimination whereby Internet retailers are able to single
out individual consumers to charge them unique prices.
Market structures may change if the Internet reduces market friction. Intermediaries, market participants who
enable transactions between suppliers and consumers, may be in an unsustainable position with the introduction of
the Internet. Intermediaries exist to coordinate transactions and reduce the overall transaction costs in market
exchanges. Unlike Adam Smith’s theory that markets existing without coordination because of an “invisible hand,”
Chandler (1977) describes why the visible hand of management is necessary to coordinate firms within markets.
Even though intermediaries are important in physical markets where market friction is significant, their role of
reducing transaction costs is threatened when suppliers and consumers use the Internet to transact. The
intermediaries who aggregate products and set prices on the Internet may be threatened if they add minimal value to
Internet transactions. The impact on market structure may be significant. Intermediaries are major contributors to
the U.S. economy. Intermediaries account for over 15% of the U.S. Gross Domestic Product (GDP) with the retail
industry accounting for 9.3% and the wholesale industry accounting for 6.5% of the GDP in 1994 (Spulber 1996). If
intermediaries disappear, then significant segments of the economy may shrink resulting in unemployment.
This thesis investigates intermediaries that participate in Internet commerce to determine how the Internet reduces
market friction and reduces the role of the intermediary. The approach of this thesis is interdisciplinary because it
incorporates economics, technology, and policy analysis. This thesis extends existing theory, develops a
methodology to analyze price competition on the Internet, and uses an exploratory data set for empirical analysis to
test the following hypothesis:
Internet commerce will reduce the market friction of physical market transactions.
Finding evidence to support or refute this hypothesis can help shape future development of Internet commerce. If
Internet commerce does reduce market friction, then there may be shifts in the global economy as consumers adopt
the Internet as a medium for commerce and intermediaries are removed from the value chain. The economic theory,
which describes changes in transaction and menu costs, will be supported by an Internet economy that has less
friction than the physical economy. However, data might also indicate that the Internet does not yet live up to the
promise of promoting frictionless markets. Transaction costs and menu costs may still be significant enough to
warrant the existence of intermediaries to coordinate transactions. The market participants must rely on strategies to
make the best use of the Internet for commerce because economic forces may not work perfectly. For example,
Internet market friction may change the shopping behavior of consumers who are trying to maximize their utility
and the retailers who are trying to set a price.
12 CHAPTER ONE
Regardless of the thesis findings, it is important to remember that the Internet is still a changing communications
infrastructure by design. Therefore, it is too early to predict what applications the Internet will support in twenty
years and who will use the Internet to conduct commerce. The businesses that are currently conducting commerce
on the Internet will use consumer feedback to change their pricing, marketing, advertising, and product offerings in
the future.2 By addressing the questions of market frictions in Internet commerce in an embryonic stage, future
analyses of Internet commerce are possible. Furthermore, the beginnings of empirical research can help identify
strengths and weaknesses of the existing theory that describes how information technology such as the Internet and
intermediaries reduce market friction.
2 Bill Rollinson of the Internet Shopping Network claimed that his business strategy was only valid for a period of a few weeks during a March
1996 presentation to the MIT Electronic Commerce/Marketing and the Internet class. Because there is no “ right way” to develop an Internet
commerce strategy, Internet Shopping Network experiments constantly and uses the feedback from consumers to help develop its strategy.
2. CONTEXT AND LITERATURE REVIEW
This chapter examines the economics of market friction and market structure, and explains them within the context
of the Internet. The introduction of a new technology does not mean that markets depart from their economic
principles. Rather, the markets adjust to absorb the new technology to a new equilibrium. The purpose of this
chapter is to provide the economic foundation and motivation for examination of the hypothesis of this thesis:
Internet commerce will reduce the market friction of physical market transactions.
The chapter defines market friction and describes the economics of the Internet. Market friction is a general
economic concept that describes the forces that shift the equilibrium of the market away from the intersection of
supply and demand (i.e. the competitive equilibrium). Section 2.1 describes market friction in a static environment
by describing the effect of transaction costs for intermediated and disintermediated market structures. Section 2.2
explores the dynamics of market friction by describing how menu costs may prohibit changes over time to reach the
competitive equilibrium. Section 2.3 describes the economics of the Internet and describes why the Internet may
reduce market friction.
2.1. Transaction Costs and Intermediation
A static examination of markets indicates that some transactions do not occur at competitive market equilibrium
because of transaction costs. Because the exchange of goods or services and monetary wealth is not free, the price
of a good that is transacted often reflects costs involved with coordination of the transaction and not the creation of
the good. For example, it may only cost $10 to manufacturer a pair of shoes in Taiwan, but it costs money to
advertise, package, and distribute the product. None of these additional costs change the physical product, but these
costs aid in finding a consumer for the shoes and receiving money for the purchase.
The examination of transaction costs began with efforts to understand why firms are created to organize economic
activity. Coase (1937) argued that firms organize themselves to minimize transaction costs so that they can be more
14 CHAPTER TWO
economically efficient.3 Others have used transaction costs to extend Coase’s work to describe why firms are
created and what distinguishes the boundary of one firm from the boundary of another (Alchian and Demsetz 1972;
Demsetz 1968; Williamson 1979). As argued by Demsetz (1968) and Williamson (1979), procuring a product can
be done within the boundary of the firm or done as a market transaction between firms. Whichever organizational
model has lower transaction costs is preferred. Transaction costs are the costs incurred when goods or services are
exchanged and not the costs associated with creating the good or service. Defining what is a firm and what is the
boundary of the firm are ongoing issues in economics. The exploration of transaction costs is only one of many
theories. For example, Hart (1989), Varian (1992), and Pindyck and Rubinfeld (1995) describes differences in the
methodologies used in neoclassical economics, transaction cost economics, and by economists analyzing the firm as
a nexus of contracts or owners of property. While some theories of the firm can give insight into choices of labor
versus technology or the cost savings of introducing one technology over another, transaction cost economics is a
more appropriate theory for understanding market friction than the alternative approaches.
Transaction costs may decrease when information technology is used to facilitate market exchanges. As
transactions become electronic, they may cost less than physical world transactions. Some argue that information
technology may lower transaction costs because of lower 1) search costs (Bakos 1997), 2) coordination costs
(Malone, Yates, and Benjamin 1987), and 3) payment processing costs (Sirbu and Tyger 1995). If transaction costs
decrease within the firm more than they decrease between firms, then there will be an organizational shift from
market transactions to intra-firm transactions. If the reverse is true, there will be more market transactions and
fewer intra-firm transactions. Because the effect of information technology on transaction costs is far from certain,
the further information technology research can explain the factors influencing the increase or reduction of
transaction costs. This section investigates the foundation of transaction costs in two market structures:
disintermediated and intermediated.
2.1.1. Transaction Costs in a Disintermediated Market
The simplest transaction costs incurred in a direct transaction between a supplier and a consumer. The supplier is
the firm that produces a product or service being exchanged and competes with other firms whose product can be a
substitute. The consumer is the end user who derives value from possessing or consuming the product. This direct
transaction does not require an outside participant (i.e. an intermediary) to coordinate the exchange between the
supplier and consumer. Therefore, this direct exchange is “disintermediated.”
The supplier, consumer, or both may incur the transaction costs. Figure 2.1 shows the affect of transaction costs
absorbed by the supplier (on the left) and by the consumer (on the right). The effect of the transaction cost is to shift
3However, the competitive equilibrium will differ for different initial allocations of goods.
CONTEXT AND LITERATURE REVIEW 15
the supply or demand curves, respectively, to the left because of higher transaction costs. Because the market is not
frictionless, either party must pay an additional amount to transact the good.
Figure 2.1: Transaction Costs with Disintermediation
S
S’
D
Q’ Q
P
P’ S
D
D’
Q’ Q
P
P’
Supplier absorbs transaction costs Consumer absorbs transaction costs
Source: Demsetz (1968)
Transaction costs related to imperfect information provide an excellent example to describe shifts in the supply and
demand curves in Figure 2.1. A supplier may have the best product on the market at the lowest cost, but they have
no sales because no consumer knows about them. Therefore, the supplier can advertise, direct market, or develop a
World Wide Web (a.k.a. Web) site for their product to attract consumers to transact with. All costs associated with
disseminating information to promote market exchanges with consumers are transaction costs absorbed by the
supplier. Conversely, consumers may want to purchase a product, but they do not know if the product exists, or who
sells it, or at what price. The consumer must search for information, communicate and negotiate with potential
suppliers before they can purchase the product. These costs borne by the consumer are also transaction costs.
The effect of the transaction cost is to decrease the quantity exchanged between supplier and consumer regardless of
who absorbs the cost. When a supplier absorbs the transaction cost, there is a class of consumers that would have
transacted at P that do not transact at P’ because the price is too high. These consumers account for the decreased
quantity of Q – Q’. Similarly, when a consumer absorbs the transaction costs, there are a class of suppliers that do
not transact because P’ is too low. Once again, there is a reduction in the quantity exchanged in this market that is Q
– Q’. The quantity of transactions that did not consummate because of transaction costs is known in economics as
deadweight loss. Deadweight loss is regarded as an undesirable outcome, to be minimized whenever possible.
16 CHAPTER TWO
The nature of the product being transacted affects the amount of a transaction cost. Williamson (1979) points out
that the transaction costs for goods with low asset specificity (such as commodities) are much lower than goods with
higher asset specificity. As an example, a New York Times bestseller has low asset specificity because it is
designed to appeal to a mass market and consumers choose to purchase or not purchase such an item at a given
price. A power plant serving a factory, on the other hand, has a high degree of asset specificity because it is
designed for a particular consumer with specific needs. There are very high production costs, and neither the
consumer nor the supplier of the power plant can exit the contract easily because the plant has very little appeal to
any other consumer. Because all people have bounded rationality,4 complex contracts for a transaction may not take
into consideration all possible events. As contracts become more complex and less complete, transaction costs
increase.
2.1.2. Transaction Costs in an Intermediated Market
Transaction costs can also be absorbed by a third party other than a consumer and a supplier—and intermediary. The
intermediary is the firm that sells the product but does not create or consume it. Therefore, intermediaries compete
with other firms who may sell the exact same product or service. Existing definitions in the economic literature
consistently define an intermediary as a firm between the supplier (producer) and consumer (buyer).5 For example,
Spulber (1996) describes an intermediary as “an economic agent that purchases from suppliers for resale to buyers
or that helps buyers and sellers meet and transact.”6 Similarly, Cosimano (1996) describes them as an institution
“between buyers and sellers.”7 Biglaiser (1993) differentiates the intermediary from the supplier and consumer by
examining its objective for participating in a market transaction. He explains that the supplier is the originator of the
good through original ownership or creation and the intermediary does not alter the good. Similarly, the
intermediary is different than the consumer because unlike the consumer, the intermediary derives no utility from
possessing or consuming the good.
The intermediary changes the transaction costs of a market transaction by buying from suppliers at one price and
selling to consumers at a different price. The stock market is one example where an intermediary, the broker, does
this. In Figure 2.2, the transaction costs are absorbed by an intermediary who has two prices—an ask price and a bid
price—for the transaction. The ask price is denoted by A in Figure 2.2 and represents the price that the intermediary
sells the good to the consumer. The bid price is denoted by B in Figure 2.2 and represents the price that the
intermediary buys the good. P is the price at the competitive market equilibrium. The difference between A and B
4 Bounded rationality describes the condition that parties cannot plan for all possible future events and they realize this when they make decisions
and enter into contracts.
5 For sake of clarity, this thesis will use consumer instead of buyer and supplier instead of producer, but they can be used int erchangeably.
6 p. 135.
7 p. 131.
CONTEXT AND LITERATURE REVIEW 17
in Figure 2.2 is the ask-bid spread and can be thought of as the transaction cost for a market exchange. Demsetz
(1968) used Coase’s explanation of transaction costs to explain the ask-bid spread in the securities market. Demsetz
explained that there is a time that a broker must hold on to the asset before they can sell it. There is a cost associated
with this time when the value of that asset may drop in price. Therefore stockbrokers buy at a price less than the
market equilibrium price ( P ) and sell at a price slightly higher than the market equilibrium price. If a market is in
greater flux, the ask-bid spread increases Furthermore, as the cost to coordinate the exchange increases, the ask-bid
spread increases.
Figure 2.2: Transaction Costs with Intermediation
S
S’
D
D’
Q
Ask (A)
Bid (B)
Transaction
cost
Q’
P
Source: Demsetz 1968
Intermediaries may be in a better position to lower transaction costs than a supplier or a consumer. Since the
intermediary is involved in many repeated transactions, they develop a set of relationships and experience that may
lower the transaction cost. Furthermore, the intermediary could invest in technology that requires a large fixed cost,
but reduces the marginal cost for additional transactions. The intermediary can then amortize the fixed cost over a
larger number of transactions.
Although intermediaries lower transaction costs, it is not clear what roles or what value they provide. The literature
agrees that an intermediary is a market participant that coordinates transactions between a consumer and supplier,
but more complex definitions detailing the roles of intermediaries are inconsistent. Often the roles of the
intermediary are context dependent because they provide different roles for different sets of transactions. Confusion
over the role of an intermediary is exemplified by the inconsistency in defining an intermediary from an economic
perspective and other perspectives such as marketing (Dwyer, Schurr, and Oh 1987) where the value of an
intermediary is very subjective. It is surprising to find a lack of consensus or attention given the intermediary’s
18 CHAPTER TWO
importance. Analysis of intermediaries is “largely ignored by the standard theoretical literature.”8 Chapter Three
extends existing theory to help define roles of intermediaries and hypothesize how the Internet may change these
roles.
There may be multiple intermediaries that separate the supplier from the consumer. Figure 2.3 visualizes the value
chain to show how a product is transacted in intermediated and disintermediated markets.9 In the top case, there are
two firms (wholesaler and retailer) that are intermediaries. In the middle case, there is only one intermediary–the
broker. In the bottom case there is no intermediary so it is called a “disintermediated” market.
Figure 2.3: Intermediated and Disintermediated Value Chains
supplier
supplier
supplier consumer
consumer
consumer
wholesaler retailer
broker
Intermediated
(one level)
Intermediated
(two levels)
Disintermediated
$50 $70 $100
$60 $90
$50
Figure 2.3 shows that the products’ price changing when an intermediary processes the transaction from a supplier
or other intermediary to a consumer or another intermediary. In the two-level intermediated case, the wholesaler has
an ask-bid spread of $20 and the retailer has a spread of $30. When the wholesaler and retailer are vertically
integrated, they form the one-level intermediated case where their spread is still $30, and both the supplier and
consumer receive an extra $10 of surplus minus their transaction costs. If the supplier and consumer transact
directly, they can transact at $50. The $50 price may seem to indicate that the disintermediated case is the most
preferred, but the consumer may have $60 of transaction costs in the disintermediated market. If consumer
transaction costs are too high, then the consumer may prefer the one-level or two-level intermediated case.
Intermediation and disintermediation describes the addition or removal of elements between the supplier and
consumer in the value chain as shown in Figure 2.3. Disintermediation occurs when an intermediary is removed
from a transaction. Intermediation occurs when one is added. The disintermediation process does not necessarily
mean that the result is a disintermediated value chain as shown in Figure 2.3. Disintermediation can also occur as
8 Rubinstein and Wolinsky (1987), p. 581.
CONTEXT AND LITERATURE REVIEW 19
the value chain shifts from n layers to n – 1 layers of intermediation. For example, disintermediation occurs when
the market moves from two levels to one level of intermediaries (moving from the top to the middle value chain in
Figure 2.3). As was noted earlier in this thesis, the popular assumption is that electronic commerce in general and
Internet commerce in particular leads to disintermediation. The theoretical basis for this view, as well as reasons
why it may or may not be true, is further developed in the remainder of this chapter and in Chapter Three.
2.2. Menu Costs and Rate of Change
Market friction can also be caused in dynamic markets through such economic factors as menu costs. The costs of
changing prices are known as menu costs. Since neoclassical economics often analyzes markets in competitive
equilibrium, it is important to understand how long it takes for markets to achieve this equilibrium. Markets with
more friction take longer to achieve competitive equilibrium while markets with less fiction take less time. When
the market is at competitive equilibrium, prices do not change. Sometimes the costs of changing prices are too high
to warrant achieving equilibrium and this increases the time a market spends in a state other than competitive
equilibrium.
Menu costs are the economic costs of “printing menus” which contain the prices for the items carried by a supplier
or intermediary. Much of the economic exploration of menu costs examines its macroeconomic effects. For
example, Sheshinski and Weiss (1993) describe how menu costs affect the setting of prices and often explain nonoptimal
prices in an economy. In a country with high inflation, it is impossible for every product price to be
changed daily even though this may be necessary for optimal pricing. Research on the microeconomics of menu
costs is not as well developed. Microeconomic studies such as Levy, et al. (1997) estimate menu costs in a given
market. In their study, Levy, et al. find that the menu costs in grocery stores is approximately $0.52 per price
change. While this may not be a significant menu costs to stores which sell high-priced items such as cars, $0.52
per price change is significant enough such that supermarkets do not often change the prices on groceries that may
be $10 or less.
When there is an exogenous force in the marketplace, prices will change more quickly or more slowly because of
menu costs. When menu costs are high, the entry of a competitor or a rise in inflation may not be significant enough
to change prices quickly. This is especially true if the response to such an exogenous force (either lowering or
increasing prices) fluctuates. However, if the menu costs are low, prices can change quickly because of exogenous
forces and fluctuate just as often as the market conditions change.
9 This value chain shown in Figure 2.3 is similar to the one used by Benjamin and Wigand (1995).
20 CHAPTER TWO
2.3. Electronic Markets and Internet Commerce
The creation of the Internet as a medium for economic activity is rooted in the technical foundations of the
infrastructure. The designers of the Internet may not have envisioned Internet commerce, but they did architect the
protocols necessary for new products and services to be developed in the future. The markets that have emerged are
overlaid on top of this technical foundation. This foundation affects the Internet’s user community, applications,
and markets. Because the Internet’s core is a minimal set of protocols which allows for heterogeneity and
development of new services, Internet commerce is possible.
The Internet infrastructure grew from the foundation of the existing communication modes. Telecommunications
has been characterized by common carriers, such as companies that operate telephone and telegraph networks,
which provides a conduit between two points. As common carriers, they must publish their prices and offer their
services non-discriminatorily (Brock 1981). For some Internet applications, the Internet appears to be a common
carrier. For example, applications such as email have similar characteristics to telegraph messages. A different
communication mode that also has similarities to the Internet is broadcasting because data can be multicast on the
Internet whereby data streams originate at one point but spread to many others. Broadcasters are regulated because
electromagnetic spectrum is scarce and, therefore, the government has an interest in making sure the content sent by
broadcasters, such as FM/AM radio and television, is consistent with the public interest. The third communication
model the Internet parallels are publishers because the Internet allows for publishing information on servers such as
Web servers. The publisher model comes from book, newspaper, and magazine publishing.
The Internet is at a point of convergence where previous communication modes come together. Because the Internet
may be many different communication modes simultaneously, the Internet is the closest manifestation of the
“convergence of modes” described by Pool (1983) and others since then (Brock 1994; Neuman, McKnight, and
Solomon 1997). The result of a larger network with greater functions increases the economic value of the Internet
(Economides 1994). Therefore the Internet has become a platform for new and novel applications including those
related to market transactions (McKnight and Bailey 1997a; 1997b). These modes are not technical categories, but
the industry and regulatory paradigms that describe different communication applications and channels. Because the
Internet spans these boundaries, it is not clear if the Internet should be regulated in the three communication modes
or whether it is a separate entity.
As telecommunication services become more liberalized, more consumers have access to the Internet infrastructure.
Internet growth and some users’ Internet access are dependent on the public switched telephone network. As the
telecommunications industry changes, so does Internet development. Agreements, such as The 1997 Group on
Basic Telecommunications agreement, liberalize the international telecommunications market. Therefore, the global
telecommunications market may become less monopolistic and more open to Internet development. However, it is
CONTEXT AND LITERATURE REVIEW 21
yet to be seen if this agreement is just confirming an existing movement or a policy which will have widespread
consequences as Drake and Noam (1997) debate.
Historically, the technology and policy of the Internet has affected its economic development. Internet development
is rooted in computer science’s quest to build a global computer network that is designed for the transportation of
digital bits through computers regardless of the appliance or application. This design philosophy of the Internet
(Clark 1988) is well rooted in technology development and not the creation of wealth that typifies other
developments such as telephony in the telecommunications industry (Brock 1981). As the technical robustness on
the network to provide simple applications progressed, computer scientists have built on the core Internet protocols
for further developments such as network security for commerce (Camp and Sirbu 1997)10 and adaptability to
congestion.
Much of the attention on the economics of the Internet has been focused on the infrastructure. As the Internet was
privatized—which occurred when the National Science Foundation’s backbone11 was dismantled—there were so
many misconceptions about the economics of the Internet that MacKie-Mason and Varian (1994) documented the
frequently asked questions of how and why the Internet market worked. Unlike circuit-switched networks, Internet
bandwidth is statistically shared so users have a dynamically allocated bandwidth regardless of the price they are
willing to pay. While this works well for asynchronous traffic such as email, statistical sharing is problematic for
services that require better qualities of service. By pricing bandwidth, Internet resource allocation changes to reflect
users’ heterogeneous demand. Wang, Peha, and Sirbu (1997) and Gupta, Stahl, and Whinston (1997) both propose
that pricing bandwidth may increase the aggregate benefits of Internet service. While these proposals are attractive
to the economist, computer scientists may find them too difficult to implement. For example, the MacKie-Mason
and Varian (1995) proposal to price each packet whereby auctions would be held at Internet routers to determine
queuing causes tremendous router overhead and would subvert the robust flow control built into the Transmission
Control Protocol (TCP). Extensions of Internet standards and technology to support a variety of service classes and
economic models is ongoing, but a more thorough review of the literature and the evolving theoretical and empirical
basis of Internet economics is beyond the cope of this thesis, which focuses on Internet commerce and the role of
intermediaries.
10 Making Internet commerce more secure increases its ability to support financial transactions. The Internet may become a mechanism to
support transactions of very small monetary magnitude called micropayments. The impact of micropayments may greatly affect the way
consumers purchase and use software. For example, Cox (1996) describes a “ superdistribution” as a way to distribute software and then use
micropayment to pay for the marginal use of that software..
11 The Internet was just a portion of an overall policy effort to build a National Information Infrastructure (NII) in the early to mid-1990s.
However, the Internet has become the focus of NII development lately. Both Kahin (1995) and Kalil (1995) point out that the initiative is and
will be funded by the private sector, but that the government can help play an important part of its development through such actions as funding
research and development and developing telecommunications policy.
22 CHAPTER TWO
Even though the Internet is a convergence of modes, the Internet enables interoperability more effectively than other
communication modes. The Internet, a “network of networks,” integrates many different data networks (Bertsekas
and Gallager 1992) into a larger network without requiring complete conformity to one standard. As Clark (1988)
explains, the Internet protocols were designed to meet the needs of communication between computers in a
distributed and heterogeneous environment.12 Because of the Internet’s design philosophy, the Internet protocols
enable interoperability. Bailey, McKnight, and Bosco (1996) show that interoperability is different than
compatibility or interconnection because interoperability enables the successful communication across
heterogeneous systems and not within a system (compatibility) or between homogenous systems (interconnection).
To accomplish the goal of interoperability, the Internet uses a minimal set of standards or “spanning layer” (Clark
1994; NRC 1994; NRC 1996; Kavassalis, Lee, and Bailey 1997), such as the Internet Protocol (IP), to support
interoperable data communication.
Increased interoperability of infrastructure elements can increase benefits for all network users because of network
externalities. A network externality (sometimes called a “network effect”) is the cost or benefit that incumbent users
get from an additional member joining the network. Katz and Shapiro (1985; 1994) outline many of these benefits.
This benefit or cost can be direct (such as the benefit from having one more person to talk with or exchange email)
or indirect (from a larger network of users encouraging greater investment in network resources). The telephone
system evolved from a network with very little or niche benefit to one of business ubiquity because it is a product
with network externalities. While businesses often perform a cost-benefit analysis when contemplating Internet
connections for their employees, this same company will most likely not undergo an analysis for a telephone system.
As the Internet’s network externalities grow, the Internet’s value may increase to such an extent that businesses will
no longer contemplate whether or not they should have an Internet connection. Rather, the Internet will become as
much of a business necessity as the telephone. As more consumers are drawn to the Internet to conduct business
because of word-of-mouth recommendations and advertising, network externalities not only increase the value of the
Internet but can also be used as a competitive tool. For example, Amazon.com solicits book reviews from their
consumers and uses this information to help other consumers with book recommendations.
The Internet’s interoperability allows for the emergence of transactions and the creation of market relationships.
Interoperability is important to the working of electronic commerce as pointed out by Petreley (1997):
“When electronic commerce becomes the standard conduit for business transactions, the dependability and
interoperability of your machines and your business-partners’ machines could mean the difference between
success and bankruptcy.”
For Internet commerce applications to function effectively, the interoperability of the spanning layer, middleware,
and applications must exist. Because the Internet was designed for interoperability, the Internet may be better suited
as a medium for electronic commerce transactions than other media.
12 One result of designing the Internet to be a distributed network is that the Internet governance structure is very self-governing (Gillett and
Kapor, 1997).
CONTEXT AND LITERATURE REVIEW 23
Electronic commerce is the set of market transactions that are facilitated by an electronic medium.13 Because there
are many electronic media and many forms of market transactions, electronic commerce is a term that is used to
describe many different forms of transactions that use information technology. Therefore it is important to
distinguish the forms of electronic commerce as suggested by Wigand (1997). There are varying degrees of
“electronic” in electronic commerce. As Choi, Stahl, and Whinston (1997) describe, electronic commerce has been
around since the beginning of electronic communication–the telegraph. Sears Roebuck was able to sell items from
its catalog and use electronic telegraph communication to receive and process orders. However, the Internet is more
advanced than this early form of electronic commerce because the Internet is interoperable while the telegraph is
slow and not very accessible. Telephone-enabled commerce aided the process of making commerce increasingly
electronic and the Internet continues this process. Figure 2.4 shows the different axes of movement of electronic
commerce: process, product, and players. At the origin is physical commerce and at the point farthest from the
origin on the cube is completely electronic commerce.
Electronic commerce initiatives stem from many technologies and computing developments. Companies such as
Books.com have used Internet applications such as telnet to conduct commerce on the Internet before the Web was
fully-developed (Gardner 1998). In France, the Minitel computer and communications network has supported
electronic commerce for more than fifteen years (Hill 1997). The use of Electronic Data Interchange (EDI) for
electronic document transfer and transaction sets was, perhaps, the most successful use of information technology in
business-to-business transactions prior to the emergence of the Internet. However, the multibillion-dollar EDI
industry is experiencing slower growth than commerce conducted over the Internet. One reason for slow EDI
growth is that EDI transactions are limited to partners who have an established set of relationships and have a highly
standardized way of communicating. Therefore, once partners have the infrastructure in place, there is less room to
solicit new consumers or to have communication outside of the a priori set of standards. The Internet, on the other
hand, does a better job of identifying the ad hoc communication channels necessary to go beyond the scope of
standardized transactions. In fact, electronic commerce that takes the EDI and Internet capabilities can be very
complementary (Kalakota and Whinston 1997a) and EDI transactions can now be sent over the Internet (ANX 1996;
Houser, Griffin, and Hage 1996).
13 Dictionaries define “ electronic” as anything relating to the conduction of electrons. An electronic medium is any conduit that transfers
electrons to convey information.
24 CHAPTER TWO
Figure 2.4: Product, Process, and Player “Degrees” of Electronic Commerce
players
process
product
completely
electronic
commerce
completely
physical
commerce
Source: Choi, Stahl, and Whinston (1997)
The Internet makes commerce more electronic on the process and players axes in Figure 2.4. The Internet can be
used to distribute electronic products such as software, music, and information. The Internet also creates electronic
players with the introduction of electronic storefronts. These storefronts often take the form of a Web page that has
product and price information. The Web site may also process and clear payment transactions. Electronic mail can
keep a consumer up to date on the status of their order just as a good salesperson would do. One of the benefits of
having electronic players in commerce is that the electronic player is infinitely patient. For example, some people
who buy airline tickets on the Web can try out a large number of permutations of flight travel routes while a human
travel agent must reduce this number to service other clients. Figure 2.5 shows the vertical distinction of electronic
commerce into Internet and non-Internet supported commerce.
Figure 2.5: Defining Internet Commerce and Electronic Markets
Internet non-Internet
Market
Hierarchy
The horizontal space defined in Figure 2.5 can be split into the two organizational parts of markets and hierarchies.
The definition of each comes from Williamson (1975) and is expanded upon by Malone, Yates, and Benjamin
CONTEXT AND LITERATURE REVIEW 25
(1987) to cases where the market exists in an electronic environment.14 A market requires little or no coordination
between transacting parties because the good being transacted is widely available from competitors. These goods
are said to have low asset specificity. Market transactions require very little information to be exchanged between
transacting parties because the consumer’s options are either to buy at a specified price or not to buy. When a
transaction goes sour, the consumer tries to find anther supplier and the supplier finds another consumer. A
hierarchy requires much more coordination. More coordination may be necessary because the supplier and
consumer transact often, or the product being transacted is unique. The prices for transactions in a hierarchy are
usually negotiated and exiting the contractual relationship of a hierarchy is often difficult. This thesis explores the
upper right-hand quadrant in Figure 2.5 where “Internet” and “Market” intersect within the electronic commerce
space. The transactions in this quadrant require little coordination because products have low asset specificity and
are transacted through the Internet.
2.3.1. The Nascent Stage of Internet Commerce
While electronic commerce on the Internet is still in its nascent stage, firms seeking to do business on the Internet
face great opportunities and equally great challenges. Internet commerce transactions at the time of this thesis’
writing are small relative to its potential. According to the Baruch College-Harris Poll (Cortese 1997) 1% of all
Internet users often shop on the Internet, 9% sometimes shop, 26% rarely shop, and 64% never shop as of late 1996.
24% of Internet users have actually made a purchase while 76% have not. These numbers are small relative to the
number of people who use the Web. There are approximately 40 million U.S. users of the Web and the Internet
(Cortese 1997).15 Because the U.S. accounts for approximately 66% of the Internet users worldwide (Economist
1997), then approximately 60.6 million people access the Web a year–a number that has been doubling every year.
The growth can be explained by an increasing number of items for sale on the Internet, more users, and technology
development. The amount of Internet sales may also be small because some users gather information on the Internet
but shop in physical stores. As a CommerceNet survey indicates, already 39% of Internet users used the Internet to
get information about a purchase while 15% of Internet users actually purchased something. As the Web and the
Internet evolves, transaction fulfillment will become more accessible to those who also find information about
products on the Internet.
Even though Internet commerce transactions were worth only $500 million to $600 million in 1996 (Economist
1997) they are forecast to increase quickly. This number will likely grow again approximately ten times this figure
14 When businesses use electronic commerce to transact, they usually require more coordination between the businesses. Therefore, businesses
in electronic hierarchies often use tools to reduce coordination costs like Electronic Data Interchange (EDI) which are very standardized. Less
standardized transaction media such as the Internet may be more suitable to markets which require less coordination.
15 According to IDC (1997), there were 27.6 million users of the Web worldwide in 1996 and 50 million in 1997. IDC (1997) estimates the
number of worldwide Web users will grow to an estimated 126 million by the year 2001.
26 CHAPTER TWO
for consumer industries by the year 2000.16 Jaffray (1997) estimates that by 2001 purchases on the Web will total
$228 billion in both the consumer and business-to-business markets. Jaffray (1997) estimates that only 11% of
Internet commerce, or $25 billion, will be in the consumer market by the year 2001 which is fifty times current
estimates. Even though this is still a small percentage of the global economy, the growth is substantial enough to be
noticed even by the largest nations of the world. It is important to note again at this time that the focus of this thesis
is on retail Internet commerce, rather than the wholesale or business-to-business transactions. Elucidation of the
early development of Internet commerce in the business-to-business sector, and the role of intermediaries in
particular in that sector is beyond the scope of this thesis.
It appears as though consumers are still experimenting with the new medium for commerce. By dividing the
number of Internet users in 1996 (60.6 million)17 by the total worth of Internet commerce transactions on the
Internet ($500 million to $600 million according to the Economist 1997), Internet users spend $8.25 to $9.90 per
year on average. If the number of Internet users does doubles by the year 2000, then 969.6 million users will
account for the $4 billion to $10 billion of commerce per year as reported by The Economist (1997). This is only
$4.13 to $10.31 per user per year. These numbers indicate that Internet users are still experimenting with the
Internet commerce transactions.
However, consumers who join the Internet commerce foray may no longer be testing the waters in the year 2000.
To them, Internet commerce becomes a viable medium for serious commercial transactions. If an estimated 60.6
million Internet users today increase their spending from $10/year to $100 per year by 2000, then over $6 billion
will be spent by existing Internet users in the year 2000. This does not include the other Internet users (969.6
million minus 60.6 million) who will undoubtedly account for a sizable percentage of Internet commerce.18 If the
remainder of the Internet users only spent $10/year, then Internet commerce would be approximately $15
billion/year by 2000 making the higher estimates seem low. In an industry that experiences double-digit annual
growth there is bound to be creativity and experimentation with business strategy (Kambil 1995; 1997). Because
Internet commerce is and will be experiencing triple-digit annual growth, this creativity and market entry is likely to
be even more prolific.
16 The actual increase is debated by different forecasters, but the most conservative, Cowies/SMBA estimates a $4 billion/year market by 2000
while Yankee Group estimate $10 billion/year (Economist 1997). Other estimates fall between these high and low values. Forrester, IDC,
Jupiter, and the Multimedia Research group estimate that electronic commerce will grow to approximately $7 billion/year in 2000. These
estimates appear to be small because the latest numbers from 1997 indicate that Internet commerce is now worth $7.8 billion (according to
discussions with Lee McKnight on February 25, 198).
17 The estimate of 60.6 million Web users is determined by extrapolating the number of U.S. Web users to the global Internet. There are
approximately 40 million U.S. users of the Web and the Internet (Cortese 1997). Since the U.S. accounts for approximately 66% of the Internet
users worldwide (Economist 1997), approximately 60.6 million people access the Web a year.
18 This section estimates almost 1 billion users by the year 2000 for the calculations. Other estimates are more conservative because they
estimate the number of Internet users to be less than half a billion.
CONTEXT AND LITERATURE REVIEW 27
The Internet marketplace may never replace the physical marketplace, but it may have far-reaching effects on it.
Because the Internet is fundamentally an infrastructure to support communication, it can be used at many different
points during an exchange in the physical world, and may introduce a variety of changes to the traditional physical
marketplace transaction. Only at the limit, when the Internet channels of communication substitute fully for
physical channels of communication, does Internet commerce threaten the existence of physical retailers. Starting
from the first principle that Internet and physical commerce will co-exist and, in many ways, be complementary and
integrated to transactions, this section explores why the Internet changes market friction.
2.3.2. Transaction Costs and the Internet
Transaction costs include the fixed costs necessary to support market exchanges. As explained earlier in this
chapter, these costs do not change the product but aid in clearing the market. In physical markets these costs range
from the property necessary to showcase inventory to consumers to the cost of advertising. While advertising costs
may still be present, retailers do not need physical locations to showcase their inventory. However, there are some
other fixed costs that are necessary for Internet retailers to support market exchanges on the Internet. The cost of
providing Web sites and Internet connections are two such costs.
The fixed cost to create an Internet presence may be lower than the costs to create a physical presence. While a
physical presence requires selection of a proper location with the facilities necessary to serve consumers, an Internet
server can be in a very remote location and even may be shared with other suppliers and intermediaries. The “look
and feel” of shopping is different, but the cost to create that atmosphere is lower. The impact on competition is a
lower fixed cost of entry that encourages market entry and competition. The result is that consumers are going to
have a greater choice of suppliers (and, therefore, they have a greater need for intermediation) and product selection.
The suppliers have the benefit of lower fixed costs that increase the average cost of products it sells. If it keeps
prices the same between physical and Internet, then it has a greater profit margin. If it lowers the price and can keep
the profit margin the same, it can increase sales.
Physical market transactions are limited by geography. Within that space Internet suppliers or intermediaries can
expect to attract a particular consumer who has a limited set of choices because search costs are high. When
markets transition from a physical environment to the Internet, there may be greater competition and greater choice
for consumers now that local monopolies are threatened by the competitors across the globe. Therefore, a business
making a decision will decide to accept more competition because the business then competes in a larger
marketplace. The physical market has a correlation between sales and geography, as seen in Figure 2.6 and
consistent with Hotelling’s (1929) “linear city”:
28 CHAPTER TWO
Figure 2.6: Correlation between Business Location and Sales
Business Location
(physical distance)
Amount of Sales
Physical
Market Electronic
Market
The difference between physical and Internet market transactions is that Internet transactions have no intrinsic
correlation between physical location and sales. In this marketplace, the distribution would more closely follow the
flat line in Figure 2.6.19 The result of the changes in economics is one of greater competition but greater scope. The
geographic monopoly is challenged by the Internet retailer who provides a reasonable alternative by taking
advantage of the breadth of their Internet presence. A larger geographic scope also enables suppliers to enter the
market who would otherwise face costs too large and a geographic scope too small to support the economics of
physical market entry.
No longer can local monopolies be sustained by geographic separation from their competitors. The best product or
service can appeal to the consumer without regard to proximity to the consumer. This, coupled with the fact that
there is increasing global competition, results in suppliers needing to use their comparative advantage to provider a
better product or service (or the same product or service at a lower price) to consumers. It is the comparative
advantage that wins consumers and not their physical location. This is especially important in markets where there
are large differences in prices of the same good. For example, the price of a compact disc in Europe is much higher
than the price in the U.S. Because U.S. suppliers and intermediaries sell at a price competitive with the physical
stores in the U.S., they have a large advantage when competing with the local European retailers. While shipping
delay and costs are important elements in the overall price difference between local and remote retailers, this price
can be amortized over many products (if an order is aggregated) and is often only a fraction of the total cost.
Therefore, it is unlikely that European retailers can maintain significantly higher prices than U.S. retailers can in
markets such as compact discs. The price for many goods may never be the same as a good bought on the Internet,
but the cost differential will decrease. With a wider reach, a more-rapid speed, and lower cost of transacting in the
Internet marketplace, there is greater competition among suppliers. The advantage that suppliers previously had–a
19 The cost of shipping and physical delivery of goods ordered via the Internet will be a factor in certain cases. For example, grocery delivery
services or theatre tickets cater to geographic areas. However, in many cases, the consumer may neither no nor care where the Internet commerce
server is physically located.
CONTEXT AND LITERATURE REVIEW 29
local environment to attract consumers–is no longer a comparative advantage. Now advantages in product or
service design will dominate proximity to the consumer as a differentiating feature of suppliers.
Geography no longer puts constraints on the ability to form relationships. As the Economist (1997a) describes, the
telecommunications industry is growing increasingly competitive, connecting more users, and providing more
services than ever. This results in a “connected world” of suppliers and consumers around the globe. The idea of a
local supplier may no longer mean geographically local, but local to the community of interest a consumer has. This
change is amplified because search tools can help the consumer help identify and find the best product for their
needs. The corollary to this consumer benefit is the benefit to the supplier and/or intermediary. The Internet
commerce supplier or intermediary can design their business to reach consumers in a global market. Because there
may only be a handful of people in a town who are potential consumers, it does not make economic sense for a
supplier or intermediary to set up shop in this town. Specifically, the service can be sustained because it has the
ability to reach a great number of potential consumers. The move towards global markets is evident in the article by
Reich (1990) that explains the difficulty in determining the origin of a product or service because the components of
this product or service increasingly come from many countries.
Once in the electronic commerce environment, the strategies developed by businesses better reflects the new
communication channel. Because relationships can be developed in an ad hoc manner using a global network of
resources a project team is no longer constrained to a corporate location or full-time employees. The rise of
consultants to perform a particular task resembles an object in computer science that is introduced into the action
plan when necessary (Cox 1996). The presence created by businesses is often reflected in how it designs its
technology. How a manager designs its Internet security and technology (Kalakota and Whinston 1996; 1997b) is
important to reach consumers. Also, the design of the Web site itself (Schwartz 1997) can also convey the business
presence and service to consumers. Because consumers cannot touch and feel the products they are buying they
must trust the image portrayed by its potential supplier. Therefore, image is everything. In fact, the use of an
intermediary can help consumers get behind the facade of the Web site and increase consumer confidence.
The impact of potentially reduced transaction costs from Internet commerce to change organizations is to be
determined but existing theory can give some insight into the question. Information technology may lead to
organizational change because of lower transaction costs is a research area with much depth. Many studies have
been done regarding the impact of some information technologies, such as the phone (Pool 1977), computer
(Ferguson and Morris 1993), and software (Brynjolfsson and Kemerer 1994). Other studies explore how
information technology effects productivity (Nohria and Eccles 1992; Brynjolfsson and Hitt 1994 and 1996; Hitt and
Brynjolfsson 1994) and business-to-business relationships (Bakos and Brynjolfsson 1993a; 1993b). Of growing
importance and an important extension to this prior work is the impact the Internet has on organizations (van
Alstyne 1996). While the motivating hypothesis of information technology lowering transaction costs and changing
organizational structure is fairly straightforward, the results from such a change are difficult to predict. It is difficult
30 CHAPTER TWO
to quantify the benefits of information technology adoption. While there is a large body of theoretical literature
supporting lower transaction costs with information technology adoption, there is little empirical work. Markus and
Robey (1988) address the question of whether information technology causes organizational change or whether
organizations change the technology. The thesis now turns to menu costs to focus on the dynamic aspects of market
friction and the emergence of Internet commerce.
2.3.3. Menu Costs and the Internet
The introduction of the Internet may reduce menu costs significantly. As the creation of “menus” becomes
electronic, the only cost associated with changing this price is the marginal cost of someone entering in a new
number. Menu costs can be even lower if an algorithm can automatically render a price. For example, a retailer
may sell a given book at 30% off the list price where the list price is an exogenous variable. If the publisher changes
the list price, the retailer does not have to change their algorithm and the price will automatically be adjusted.
Reduced menu costs can also lead to phenomena such as price discrimination, for example, which is described in
Chapter Six.
Rapidly changing markets and instantaneous consumer feedback is a feature of Internet commerce. This presents
challenges for firms as they generate business plans and try to develop relationships with their consumers. Rapid
changes in Internet markets mean that today’s business plan may be obsolete tomorrow. For example, the feedback
received from consumers for newspaper and magazine ads is usually measured in days, weeks, or months. Internet
advertising can change in minutes or seconds because the responses from consumers can be processed almost
instantaneously and information can be dynamically changed. As Iansiti and MacCormack (1997) point out, the
Internet allows for fast consumer feedback that allows companies with flexible product development processes to
respond to consumer needs very quickly. In microeconomics terms, the “short run” menu cost for Internet
commerce is much shorter than the “short run” menu cost for physical markets.
The acceleration of feedback and ability to dynamically change offerings and appearance also affects competition.
The ability for a supplier or intermediary to react and respond to their competitor is much faster. Once an offering is
made over the Internet it is available for consumers and competitor alike. Because the competitors can dynamically
change, they can match their competitor’s offering. For example, a price change by a competitor can be
immediately recognized and a response can happen quickly because of the Internet’s reduced menu costs.
In the electronic commerce environment, business strategies can be more flexible and modular. For instance,
relationships can be developed in an ad hoc manner using a global network of resources. In some ways, the rise of
consultants to perform a particular task resembles the object-oriented approach to software development (Cox 1996).
The result of flexible and modular business strategies is that products and services that can be tailored to the needs
CONTEXT AND LITERATURE REVIEW 31
of individual consumers through supplier-consumer relationships. The result may be products that are “mass
customized” to the individual needs of the consumer as described by Pine, Peppers, and Rogers (1995).20
Likewise, instantaneous consumer feedback may change how firms develop relationships with their consumers.
Information technology holds promise to enrich the bonds of communication between buyer and seller (Bakos and
Brynjolfsson 1993a), and some firms are already taking advantage of this capability. This relationship development
is a growing concern in the marketing literature (Dwyer, Schurr, and Oh 1987) and marketing on the Internet.
Information technology such as the Internet changes marketing strategies to stress quality of consumer interaction
(Hoffman and Novak 1996). The introduction of information technology to the market may increase the number of
relationships and reduce consumer loyalty because new marketplaces can develop or it may strengthen the bonds of
existing relationships. In the review of the literature, Steinfield, Kraut, and Plummer (1995) discover that the latter
is more likely. Bakos and Brynjolfsson (1993a; 1993b) find that information technology increases the bonds of
communication and can bring the buyer and seller in closer contact so their relationship strengthens over time. The
ability to develop communities of interest and trust in relationships is part of the goal of Internet retailers to attract
new sales and promote repeated sales (Hodges 1997).
The following chapters explore the ability of the Internet to reduce market friction and determine what impact
possible reduced market friction has on markets. It does so by exploring the roles of intermediaries, introducing a
methodology to analyze Internet price competition, and the presentation of results from an exploratory empirical
analysis. An interdisciplinary approach is used to integrate the issues of Internet technology, economics, and policy.
Interdisciplinary analysis is not a new approach to academic analysis, but is becoming increasingly important. As de
Neufville (1988) points out, decisions are often laden with the interrelationships between technology, management,
and policy and cannot be solved independently. Therefore, interdisciplinary methodologies try to examine decisions
and model the interrelationships of variables to ensure a balance of interests are met. This will, hopefully, ensure
that good technology developments are not implemented for its inability to address business and/or policy concerns.
As Clark (1994) points out with regard to network layers and standards, markets cannot explore what technology has
precluded.
20 Hagel and Armstrong (1997) describe how mass customization concepts can be used in electronic media.

3. INTERMEDIARY ROLES AND THE INTERNET
This chapter explores intermediation in Internet commerce. While an Internet retailer participates in market
transactions to provide a variety of intermediary roles, this chapter describes the aggregation and pricing roles of the
intermediary and examines how the Internet may enhance this role. Section 3.1 builds a framework for
intermediation to identify four roles of intermediaries: aggregation, pricing, search, and trust. Section 3.2 develops
an analytical model to describe what market structures (disintermediated, markets, or hierarchies) will minimize
transaction costs. Finally, Section 3.3 uses the model and examples to describe how the Internet marketplace may
take on these different market structures and how the Internet affects intermediary roles.
3.1. Intermediary Roles
The contribution of this section to the thesis is the identification of four intermediary roles: aggregation, pricing,
search, and trust. This section accomplishes the identification of the four roles by examining existing economic
literature on intermediaries, creating an exhaustive list of intermediary functions identified by this literature, and
then grouping these functions into the four roles. Then, this section explores the process by which a consumer
purchases an item and explains why the four roles are appropriate. The discussion distinguishes each role from the
others and finds a lack of depth in exploring the aggregation and pricing roles of an intermediary. The discussion is
not specific to the Internet and Internet commerce.
The roles of the intermediaries identified in this section explore the role of the intermediary as a coordinator of
information only. As with the economic literature and the consumer search process described by Bakos (1997),
Diamond (1985), and others, this analysis will not explore the logistics and transportation elements of a transaction.
For a transaction to be fully completed, the transfer of money for goods is necessary. For example, this analysis will
not explore the role of the intermediary to support payments on the Internet which is examined elsewhere (Cox
1996; and Sirbu and Tygar 1995). Some elements of the logistics and transportation process may be integrated with
the flow of information that leads to the consumer and supplier’s decision to transact.
Grouping the roles of intermediaries as identified by existing literature on intermediation results in the four roles in
Table 3.1 Even though there may be acknowledgment of the other roles, the depth to which these roles are
34 CHAPTER THREE
examined varies from article to article. By taking the literature as a whole, the articles of varying depth complement
each other to form a more complete composite of intermediary roles. Table 3.1 lists ten articles that describe
functions of intermediaries and how these functions map into the four roles. This subsection describes the
characteristics of each of the four roles to distinguish them from each other. This discussion serves as a summary of
the existing literature and describes how Table 3.1 was derived. This material is not original, but the grouping of the
material into the four roles is. Section 3.2 develops an analytical model to describe when intermediaries are
preferred in market structures. Finally, since Sections 3.1 and 3.2 are not specific to the Internet, Section 3.3
discusses how the four intermediary roles vary with the introduction of the Internet.
INTERMEDIARY ROLES AND THE INTERNET 35
Table 3.1: Identified Roles of Intermediation
Article Aggregation Pricing Search Trust
Croson (1995) economies of scale and
scope
expert screeners delegated monitoring
reputation store
Resnick, Zeckhauser,
and Avery (1995)
pricing inefficiencies search costs
incomplete information
lack of privacy
contracting risk
Sarkar, Butler, and
Steinfield (1995)21
transaction economies
of scale
purchase influence22 search and evaluation
needs assessment and
product matching
position of consumer
information
integration of consumer
and producer needs23
consumer risk
management
producer risk
management
Bailey and Bakos
(1997)
aggregation matching
facilitation
trust
Gehrig (1993) search
Biglaiser (1993) experts
Biglaiser and Friedman
(1994)
quality
Lu (1997) Pricing Indexing and Searching
Filtering
Matching
Identity Verifying
Quality Guaranteeing
Cosimano (1996) Matching
Spulber (1996) Price Setting and
Market Clearing
Matching and Searching
Providing Liquidity and
Immediacy
Guaranteeing and
Monitoring
Note: Boldface indicates that the role is explored in greater depth in this article.
The label the intermediary has usually connotes its role in the marketplace. For example, intermediaries are often
referred to as middlemen (Rubinstein and Wolinsky 1987; Lu 1997a; Lu 1997b; Biglaiser 1993; and Yavas 1992).24
Yavas (1992) describes some intermediaries as marketmakers or matchmakers depending upon the intermediary’s
role in a transaction. A marketmaker searches for places to intermediate exchanges whereas an intermediary acts as
a matchmaker when consumers or suppliers ask them to find a suitable partner. Sarkar, Butler and Steinfield (1995;
21 Sarkar, Butler, and Steinfield (1995) also list product distribution as an intermediary role, but this thesis will not look at the product
distribution and only at the information leading up to a contract.
22 Sarkar, Butler, and Steinfield (1995) describe “ purchase influence” as the ability to market a product through product placement and pricing.
This is listed as part of an intermediary’s pricing role because the intermediary chooses the product’s price.
23 It is difficult to characterize this role because this intermediary role involves conveying information to the needs of the consumer and the
supplier. Because it reduces the time needed to process and filter information for the different market players, it is a tool to help matching and
search.
36 CHAPTER THREE
1996) use the term cybermediaries to distinguish intermediation in electronic markets from those in physical
markets. Finally, when an intermediary does not take ownership of the good transacted, the term broker (Resnick,
Zeckhauser, and Avery 1996) or agent is used.
3.1.1. Aggregation
Intermediaries can aggregate products among suppliers to reduce transaction costs. Instead of a market where each
consumer has to negotiate individually with an appropriate supplier, and each supplier will have to negotiate terms
and fill the orders of individual consumers, the intermediary can aggregate the demand of many consumers or the
products of many suppliers. The intermediary is able to offer products that originate from different suppliers that
may be substitutes or complements. For example, if the intermediary is a convenience store, it would be important
for them to have food and beverages–compliments. If the intermediary is an electronics “superstore,” they can
provide a wide selection of televisions that the consumer can choose from–substitutes. This helps consumers
compare prices and product features by aggregating products in one location. The ability of the intermediary to have
product information in one location allows for a side-by-side comparison of product features.
Aggregation of products can help intermediaries realize economies of scale and scope. By having more of the same
product, the economies of scale will share the fixed cost of transacting over a greater number of items thereby
reducing the average cost. By having a greater variety of products, the fixed costs can be spread over different
products. Economies of scale and scope are identified as an intermediary role through the work of Demsetz (1968)
and Resnick, Zeckhauser, and Avery (1995), for example. This is one reason why some retailers choose to carry a
very large selection of products–the marginal cost of carrying an additional product is small. For example,
Buxmann, Rose, and Konig (1997) describe how an aggregation intermediary is important in software exchange. In
their paper, they describe a repository for Java software elements so that many developers can store their software
code on one server as opposed to a distributed Web of servers. The benefit to consumers of the software elements is
that they only need to visit one server to get a bundle of elements previously stored on multiple servers.
Aggregation of products can also lead to setting a price for a bundle of products. An example of bundled
information is a newspaper or a magazine, which takes information from different authors and bundles it together
and sells it at one price. Digital information goods, such as a news article, a digital image or a song, allow perfect
copies to be created and distributed almost without cost via the Internet. Cable TV is also an example of digital
information where the Cable TV provider bundles programming. As shown by Bakos and Brynjolfsson (1997), a
strategy of selling a bundle of many distinct information goods for a single price often yields higher profits and
greater efficiency than selling the same goods separately. In other words, the nature of information goods and the
24 It appears from reading the articles that intermediaries and middlemen are used fairly interchangeably so the fact that they are not
distinguishable may not change their analysis or conclusions.
INTERMEDIARY ROLES AND THE INTERNET 37
emergence of a market that allows their efficient distribution, create new roles for content aggregating
intermediaries that will bundle large numbers of information goods. In markets other than information goods,
bundling to set one price can be important as seen in computer sales. Computers are often sold with software
packages and peripherals to make the product more valuable to the consumer.
Intermediaries can also aggregate consumers. When an intermediary, such as a retailer, is dependent upon one
supplier for its goods, the supplier can be a very tough bargainer when setting prices and delivery dates, for
example. However, once this intermediary carries products from the supplier’s competitors, the supplier no longer
has this bargaining power over the intermediary because the intermediary has aggregated products over a larger
number of suppliers and is not depending on any one supplier. This is similar to the vulnerability cost identified by
Malone, Yates, and Benjamin (1987) where the inability to complete a transaction between a supplier and an
intermediary does not leave the intermediary in a more vulnerable position when the intermediary aggregates
products from other suppliers a well. Aggregating consumers reduces the bargaining asymmetry of disintermediated
markets as examined by Williamson (1975).
3.1.2. Pricing
The intermediary provides an important role in setting a price for a good or service because it is the price that helps
clear markets. While the microeconomic theory indicates that price should be set where supply equals demand, the
ability to measure supply and demand in an actual market is a difficult task. Therefore, the intermediary determines
a price for a good based upon the perceived demand and the supply for a good. If the price is set too high, the
intermediary may be left with a large inventory of unsold products. If the price is set too low, products might not be
available after they sell out to the first consumers and the intermediary looses revenue. By setting a price, the
intermediary acts as a Walrasian auctioneer–an important function in microeconomics of setting a price so that a
market equilibrium can be achieved. Of course the intermediary does not set prices in a vacuum. The intermediary
is able to examine what its competitors are doing and it can track sales to determine if it should change prices over
time.
When an intermediary determines a price, they can manage the balance of producer and consumer surplus in a
transaction. The sum of this surplus is the overall social welfare in a transaction. Social welfare is increased when
there are more transactions between parties and either the supplier, consumer, or both have positive surplus. The
reason some markets do not have these transactions is the transaction cost (Coase 1937; Williamson 1979)
associated with finding information and parties to transact with. The role of intermediaries to match buyers and
sellers is often addressed by the literature on electronic markets (Malone, Yates, and Benjamin 1987; Resnick,
Zeckhauser, and Avery 1995) and intermediation (Cosimano 1996; Biglaiser 1993); and Yavas 1992).
38 CHAPTER THREE
The intermediary can also become an agent for price discrimination. Price discrimination is a tool to set different
prices for the same (or similar) good to separate users. By setting a higher price when the product is sold to
consumers who are willing to pay more, the intermediary is able to realize a greater profit. Also, by giving
discounts to other users (through coupons or student discounts, for example) they can charge a lower price for other
consumers who are not willing to pay more in order to increase sales. Microeconomics views price discrimination
as a tool to reduce consumer surplus and increase producer surplus. Because the intermediary is able to set prices
and aggregate products, they can perform a variety of price and product bundles to appeal to these different classes
of users.
3.1.3. Search
Because the intermediaries are repositories of information, the intermediary can make information more accessible
to consumers thereby reducing consumer search costs. Information transfer between organizations is costly,
especially when it involves “implicit” or contextual knowledge that cannot be easily articulated. In these instances,
an intermediary can facilitate the exchange of information by coordinating the process and translating the
information that is sent between the supplier and the consumer. More than just having information, the intermediary
can also process and handle this information to make it accessible to the consumers and suppliers. For example,
consumers can use the Internet to schedule travel plans, however many still use travel agents because travel agents
can help organize and filter information. They do so by providing information on several travel alternatives, as well
as processing payments, and printing tickets, invoices, and itineraries. Intermediaries can help prevent
misinformation from being gathered by the consumer. Because the intermediary is a repository of information
(Croson 1995) and receives feedback from consumers and suppliers about this information, the intermediary can
insure that the information it stores is accurate. The intermediary is more objective than the supplier is especially
when the intermediary aggregates products from across suppliers. The comparison shopping at an intermediary
includes this assessment of product features for products manufactured by multiple suppliers.
The search intermediary can help consumers find a product that is best suited to the consumer’s preferences. There
are two types of search for consumers: price search and features search. Both types of search require gathering and
processing information but address different aspects of their shopping experience. When looking for the best
product, a consumer performs a features search while a search for the best price is a price search. Search
intermediaries help with both types of searches. For example, one type of search intermediary is a direct marketing
firm. A direct marketer collects information about consumers and tries to determine to which set of consumers they
should target their advertising in order to maximize the potential for sales. Because new supplier entrants have no
interactions with the consumers they are trying to reach, they depend on an intermediary or another supplier to get
consumer information. Because another supplier is probably a competitor, it is unlikely that they are willing to
share consumer information with them. However, the intermediary would not only be more inclined to share this
INTERMEDIARY ROLES AND THE INTERNET 39
information, they may have a richer information source because their data can span markets and supplier
transactions.
Information asymmetries are one reason why an intermediary can help with search and matching similarly to the
way an arbiter works with parties trying to negotiate. The intermediary can know information about the consumer
and the supplier that neither knows about each other–an information asymmetry. The intermediary is able to take
advantage of information revealed by the consumer and the supplier to enhance the value of the transaction without
revealing the information to the other party. By keeping information to themselves, the intermediaries are able to
filter through the unnecessary consumer or supplier information and only reveal what is important to making that
party satisfied with the transaction.
3.1.4. Trust
Opportunistic behavior by suppliers and buyers can be monitored and prevented by an intermediary. In neoclassical
economics, suppliers and consumers are maximizing their utility even if that means reducing the utility of the other
party. While this often happens in negotiations for a mutually agreeable outcome, either the consumer or supplier
may supply misinformation to increase their utility even more. The trust role taken on by some intermediaries is
there to protect consumers and seller from the opportunistic behavior of other participants in a market (Williamson
1975). Because of their long-term participation in the market, intermediaries have high incentives to ensure that
market transactions are completed, and that each party involved–the supplier and consumer–lives up to their end of
the bargain. Because the parties in a transaction may need to interact with the intermediary in the future even if the
supplier and consumer never do business with each other again, the intermediary may be in a better position to
prevent opportunistic behavior, compared to other market participants.
The repeated nature of intermediary relationships makes them more trusted. By having an intermediary in the
middle of an exchange, the intermediaries are able to witness the actions of suppliers and consumer. With the
repeated exchange through these intermediaries, Croson (1995) argues that these intermediaries can help determine
future performance based on past. Because the intermediary is trying to develop a reputation, it avoids selling
flawed products sold by suppliers. Croson (1995) further explains that protecting ones reputation is far more
important when the repetition of transactions is fundamental to future profits. Because of this, the intermediary will
be more concerned about its reputation than a supplier and therefore promotes trust.
Intermediaries act as trusted third parties by assessing suppliers and products and granting a seal of quality to some.
For example, Consumer Reports is a trusted third party for many consumer goods because it provides an objective
analysis of products using metrics that many consumers are comfortable with. In the Internet, a non-profit
organization TRUSTe (www.truste.org) gives such seals of approval to Web sites that have proven they protect a
40 CHAPTER THREE
consumer’s privacy. Other trusted third parties help rate content on the Internet such as the Recreational Software
Advisory Council.
The trust intermediary can help prevent the market for lemons. Ackerlof (1970) describes the ability of suppliers to
produce flawed products or “lemons” and find a suitable market for them. However, the intermediary can reduce
the sale of lemons as Lu (1997a) points out because it is trying to protect its reputation as an intermediary and counts
on its reputation when competing with other intermediaries. Furthermore, because the intermediary oversees a
greater volume of sales, consumer complaints to the intermediary can influence the information the intermediary
gives consumers gathering information. Clemons and Weber (1997) argue that the introduction of information
technology will reduce transaction costs for the intermediary and, therefore, the intermediary assumes a greater role
as risk-manager.
3.2. Intermediaries in Markets and Hierarchies
This section introduces a model that merges consumer search costs and market structure theory. This model
describes market structures in order to describe how intermediaries can reduce transaction costs. The goal of the
model is to describe why some market structures are preferred when the intermediary provides an aggregation and
pricing role. The insight of the analytical model is the realization that disintermediated markets work well with very
few suppliers, hierarchies work well with a large number of suppliers, and markets work well for cases in-between.
While the model is not Internet specific, the market structure theory it builds on comes from a discussion of
electronic markets and not physical markets (Malone and Smith 1988; Malone, Yates, and Benjamin 1988; and
Bakos 1997).
Similar to the work of Cosimano (1996), this section focuses on the choice between disintermediated and
intermediated markets. As the previous section describes, these two markets underlie many debates about how the
value chain changes with the introduction of markets. Unlike Cosimano (1996), the model presented here will
address the aggregation and pricing intermediary and not the matching and search intermediary. Furthermore, the
approach taken by the model introduced here is rooted in transaction cost economics and not game theory.
Estimation of the transaction cost involves enumerating the channels of communication between all parties involved
in the market structure. Similar to the methods of measuring transaction costs used by Malone and Smith (1988)
and Baligh and Richartz25 (1967), the model introduced in this section assumes a fixed transaction cost for each
communication channel. As the number of communication channels increases, transaction costs increase. For
example, if a consumer decides to search two retailers to compare prices and product features, then they have double
25 Baligh and Richartz (1967) described the number of relationships being reduced from m * n to m + n when a single intermediary was
introduced to coordinate market transactions.
INTERMEDIARY ROLES AND THE INTERNET 41
the transaction costs of a consumer who only searches one retailer. In this example, the former consumer has two
communication channels while the latter has only one communication channel. Addition of suppliers, consumers,
and intermediaries to the market structure may increase or decrease the number of communication channels and,
therefore, change the transaction costs.
Some of the communication channels are developed when consumers search for products and features to determine
what to purchase and where to purchase it. By approaching the same problem from the consumer perspective these
two functions are not completely separate because