13-Aug-2001&

Business-to-business commerce includes a broad range of intercompany
transactions, including wholesale trade as well as company purchases of
services, resources, technology, manufactured parts and components, and
capital equipment. It also includes some types of financial transactions between
companies, such as insurance, commercial credit, bonds, securities and other
financial assets. The popular phrase “B2B e-commerce” refers to the substitution of
computer data processing and Internet communications for labor services in
the production of economic transactions. Many companies engaged in B2B
e-commerce are intermediaries between other companies that buy and sell goods
and services.
The potential size of B2B e-commerce in the economy is vast, but difficult to
pin down. Jupiter Communications (2000) estimates that overall transactions of
goods (excluding services) between businesses in the United States should amount
to $11.5 trillion in 2000, of which $336 billion are conducted electronically. (Note
that total transactions are revenue measures that may well exceed GDP, which is a
measure of total value added.) By 2005, Jupiter expects the on-line component to
represent $6.3 trillion out of a total of $15.1 trillion. A bit more modestly, Goldman
Sachs (2000) projects B2B e-commerce transactions to reach $4.5 trillion worldwide
by 2005. The Gartner Group estimates that there was $90 billion in Internet B2B
transactions in 1999, by comparison with only $16.7 billion in Internet business-toy David Lucking-Reiley is Assistant Professor of Economics, Vanderbilt University, Nashville,
Tennessee. Daniel F. Spulber is Elinor Hobbs Distinguished Professor of International
Business, Kellogg Graduate School of Management and Professor of Law, School of Law,
Northwestern University, Evanston, Illinois. Their email addresses are ^reiley@vanderbilt.edu&
and ^jems@nwu.edu&, respectively.
Journal of Economic Perspectives—Volume 15, Number 1—Winter 2001—Pages 55–68
consumer transactions, including brokerage fees for on-line financial trading as
well as retail sales of goods (Uchitelle, 2000).
Expectations about productivity gains from business-to-business e-commerce
can be usefully divided into four areas: efficiencies from automation of transactions, economic advantages of new market intermediaries, consolidation of demand and supply through organized exchanges, and changes in the extent of
vertical integration of companies.
Cost Efficiencies from Automation of Transactions
Traditionally, interbusiness transactions begin with a buyer looking for inputs
or a supplier seeking buyers for its goods and services. Buyers and suppliers search
for each other through advertising, trade shows, brokers, and dealers. Suppliers
send out sales agents. Buyers then negotiate with potential sellers concerning
product specifications and prices, and perhaps conclude a spot transaction or form
a long-term contract. After the agreement has been reached, the transaction still
involves ordering, billing, arrangements for transportation, confirmation of payments, and acceptance of delivery.
E-commerce innovations aim to reduce the cost of procurement before, during and after the transaction. At every stage, e-commerce avoids the need to
translate computer files into paper documents, a process that generally involves
errors, delay and costly clerical personnel. E-commerce automates this process by
mediating transactions through websites and electronic data interchange.1
Before the transaction, Internet technology may lower the cost of searching for
suppliers or buyers and making price and product comparisons. Search costs can be
significant relative to the value of the product, particularly for small purchases. Alf
Sherk, the founder of e-Chemicals, claims: “When you’re dealing with one or two
drum quantities, the cost of comparison shopping can be more than the value of
the product” (Jones, 1999). Sales representatives have traditionally carried out such
mundane tasks as tracking product availability and pricing, and supplying such
information to customers. By automating these information services, e-commerce
relieves sales personnel of these tasks, allowing them to concentrate on account
management and marketing strategy (Slade, 2000).
During the transaction, e-commerce can reduce the cost of communicating
1 Electronic data interchange (EDI) is an e-commerce technology older than the World Wide Web that
involves point-to-point communications done over proprietary networks, rather than over the Internet.
Relative to EDI, Internet commerce offers considerable advantages in terms of cost and convenience.
Internet commerce typically makes use of open standards and off-the-shelf technology on a global
network, while EDI relies on customized hardware and software. However, despite the relative advantages of Internet commerce, the installed base of EDI connections will likely coexist with the Internet
for some years into the future (Eriksen, 2000, p. 14). According to the U.S. Department of Commerce,
EDI is currently used at more than 250,000 companies in the United States, processing an estimated $3
trillion in transactions in 2000 (Phillips and Meeker, 2000, p. 25).
56 Journal of Economic Perspectives
with counterparts in other companies regarding transaction details. Transactions
over computer networks avoid many of the associated costs of interpersonal economic exchange, including the costs of travel, time spent on communication,
physical space for meetings, and processing paper documents.
After the transaction, electronic commerce allows companies to lower costs of
communication, to monitor contractual performance, or to confirm delivery. In
addition, companies can apply information generated by the transaction to update
their inventory, production and accounting records by automatically linking their
transactions to software used for managing all aspects of the firm including sales,
purchasing and operations.
The potential cost savings in this area are substantial. Processing a purchase
order manually, including paperwork, data entry, phone calls, faxes and approval
requests, can be quite expensive, so on-line transactions might easily reduce costs
by a factor of five or ten or more. There is anecdotal evidence that such cost
reductions are possible. British Telecom estimates that by moving external procurement functions to electronic commerce it has reduced its costs from $113 to $8
per transaction (Phillips and Meeker, 2000, p. 31). MasterCard estimates that the
cost of processing purchase orders has fallen from $125 to $40, with the time
involved cut from 4 days to 1.25 days (Alaniz and Roberts, 1999, p. 13). Lehman
Brothers finds that a financial transaction is $1.27 for a teller, $0.27 for an ATM,
and $0.01 for an on-line transaction (“Internet Economics: A Thinker’s Guide,”
2000). On-line brokerage fees have fallen to below $5 in comparison with traditional discount brokerage fees exceeding $50, suggesting a decrease in costs in
back-office operations and brokerage transactions with financial exchanges. Even if
such estimated savings are greater than average or vary across industries, their
aggregate impact is likely to be enormous. By lowering the costs of transactions,
e-commerce will change not only operating costs but also the characteristics and
scope of feasible transactions.
Economic Efficiency Gains from Intermediation in B2B
E-Commerce
Intermediation and market-making are central activities in a market-oriented
economy, bringing buyers and sellers together. Intermediaries can reduce transaction costs relative to direct exchange by reducing the costs of search, certifying
product quality, mitigating communication costs, and providing guarantees for
buyer or seller commitments (Spulber, 1996, 1999). Companies acting as marketmakers enhance transaction efficiency by creating institutions of exchange, adjusting and communicating prices, clearing markets, allocating goods, and providing
liquidity and immediacy (Spulber, 1998).
Business-to-business e-commerce appears likely to transform the traditional
patterns of intermediation. Intermediaries reduce search costs by consolidating
David Lucking-Reiley and Daniel F. Spulber 57
markets, providing market information and offering an assortment of goods and
services, so that buyers obtain the cost efficiency of one-stop shopping, rather than
spending time contacting multiple suppliers. Many business-to-business intermediaries seek to offer a broader range of services, including price adjustment and
communication of price information. Centralized markets often reduce time costs
by replacing bilateral negotiation with formal bidding mechanisms and information about transaction prices.
B2B e-commerce companies cover a wide spectrum of industries. Companies
have been formed to trade items from advertising to almonds, from lighting
fixtures to laboratory equipment, from cattle embryos to circuit boards. On-line
markets have been established (or at least announced) for advertising, aircraft
parts, agriculture, apparel, automotive parts, chemicals, computers and electronics,
energy, financial instruments, food and beverages, health care, intellectual property, freelance services, laboratory supplies, industrial machinery, metals, office
supplies, plastics, paper, printing services, shipping, telecommunications, and
travel services. One market receiving particular attention has been products for
maintenance, repair, and operations (MRO) because materials such as belts,
pumps, and light fixtures are required by almost all firms in all industries, and
therefore are seen as a large potential market.2
Many business-to-business companies hope to create additional intermediation
services. Some B2B companies propose to design innovative procurement transactions between a company and its many suppliers. Other firms are setting up markets
that can provide novel types of transactions that aggregate supply and demand. For
example, the oil companies BPAmoco, Royal Dutch/Shell Group and Totalfina Elf
Group, along with financial services companies Deutsche Bank, Goldman Sachs,
Morgan Stanley Dean Witter and Societe´ Generale, announced a new global
electronic marketplace for over-the-counter energy, metal, and other commodities
called the Intercontinental Exchange to replace transactions that were largely
conducted bilaterally by telephone (Shmukler, 2000). The Intercontinental Exchange market as envisioned will have various novel features including the provision of information to traders regarding the creditworthiness and other characteristics of their counterparties, along with information regarding market aggregates
(Market News Publishing, 2000).
A number of on-line companies have described ambitious plans for offering
some of the value-added services frequently offered by established dealers: trade
credit, supply chain management services, appraisal, transportation, storage and
other wholesale activities. For example, ChemConnect, an on-line exchange for
2 Industry insiders typically refer to a marketplace targeting a single industry, such as for steel or paper,
as a vertical market, while a marketplace targeting many industries, such as for maintenance, retail and
operations products or for shipping services, as a horizontal market. For example, VerticalNet has made
a name for itself on Wall Street by developing trading communities in a variety of different verticals,
including sites such as Bakery Online, E-Hospitality.com, Machine Tools Online, and SolidWaste.com.
We do not find this distinction terribly helpful from an economic point of view, but understanding it
helps in deciphering industry reports.
58 Journal of Economic Perspectives
chemicals and plastics, arranged for Optimum Logistics to provide transportation
services. VerticalNet provides consulting services to complement its intermediation
activities on the dozens of specialized websites it operates in different markets.
VerticalNet’s websites also feature services often referred to as “community and
content,” including industry news, reviews, and other editorial features.
Popular discussions often suggest that efficiencies in B2B e-commerce are
obtained by disintermediation: that is, by “cutting out the middlemen” and supplanting presumably costly intermediaries with direct transactions. However, less
expensive intermediation and lower transactions costs do not necessarily mean
fewer intermediaries. If B2B e-commerce encourages outsourcing to replace some
transactions previously internal to the firm, or if firms can use intermediaries to
outsource some of their current external purchasing and sales efforts, or if firms
can employ specialized intermediaries to avoid inefficient one-on-one direct meetings between companies and their suppliers, the end result would be a greater
number of intermediaries (Spulber, 1999).
E-commerce intermediaries can be classified into four main categories: brokers, auctioneers, dealers, and exchanges. These categories depend on the intermediary’s pricing mechanism and whether or not the intermediary takes ownership
of the goods and services.
Brokers match buyers and sellers for a fee. Some brokers offer referral services
that resemble yellow-page directories, but with more comprehensive information
and search facilities. Buyers generally do not pay for access, using the directories to
contact sellers who in turn pay listing fees. Examples include Buzzsaw (construction) and Bakery Online (bakery supplies).3 Sellers also can place product listings
that resemble classified ads. At the websites of some B2B brokers known as “catalog
aggregators,” buyers can view sets of catalogs that contain information about
products and prices from different sellers. Examples are iProcure (office supplies)
and e-Chemicals (industrial chemicals). Catalog aggregators take orders for products on the sellers’ behalf, typically leaving order fulfillment up to sellers.
Auctioneers take a more active role in transactions by setting up a mechanism
to determine prices. Internet technology significantly reduces the cost of running
auctions. Auctions are attractive by comparison with posted prices when there is
enough uncertainty about market-clearing prices.4 Some auctioneers hold auctions
of surplus inventory for sellers; examples include MetalSite (steel) and One Media
Place (advertising space). Other auctioneers hold reverse auctions for buyers, in
which sellers compete against each other for a procurement contract. FreeMarkets
Online, founded in 1995, is perhaps the largest “reverse auctioneer;” its 1999
3 Though these two sites provided only referral services at the time we visited them, both had plans to
add additional services to facilitate transactions. This is a general trend among B2B e-commerce
markets; most advertise plans to provide many more services than they currently have available. Building
up a user base on a “referral” site may well prove to be a good strategy for eventually creating a market
where transactions take place explicitly.
4 Auctions have also become popular among individual consumers on the Internet; see Lucking-Reiley
(2000).
Business-to-Business Electronic Commerce 59
financial statements report over $2.7 billion in transactions and $20.9 million in
corporate revenues. FreeMarkets has conducted procurement auctions for dozens
of clients, including Quaker Oats, Deere & Co., and the Pennsylvania Department
of Transportation, with bidding from over 4,000 vendors from more than 50
countries in over 70 categories (Gupta, 2000).
Dealers take ownership of goods provided by suppliers and resell them to
buyers. Dealers post ask prices for buyers and bid prices for sellers. They earn
returns from the bid-ask spread, adjusting prices to changes in market conditions.
In the early stage of development of B2B e-commerce, few of the new on-line
companies attempted to become dealers. Instead of building the physical infrastructure required for inventory and shipping, most chose to focus on building the
electronic infrastructure of markets (broker, auction or exchange).
On-line dealer intermediaries tend to be subsidiaries of “old economy” distributors with existing physical infrastructure. Established distributors have certain
advantages, including industry expertise and customer and supplier contacts.
W.W. Grainger takes advantage of its long-standing distribution network. Grainger,
a distributor of supplies for maintenance, repair, and operations founded in 1927,
makes its catalog of hundreds of thousands of items available for on-line ordering.
Interestingly, Grainger negotiates customer-specific contracts, so that each customer views a customized set of prices after logging in to Grainger.com. EnronOnline is an on-line extension of Enron’s business of buying and selling contracts for
natural gas and other commodities with over-the-counter brokers, other wholesale
merchants, and its own sales and marketing unit. Enron’s trading has migrated
from dealing over the phone to on-line buying and selling at posted prices.
According to Enron’s quarterly report for the third quarter of 2000, EnronOnline
had conducted 350,000 transactions worth $183 billion since its launch in November 1999.
Exchanges are double-sided markets, similar to existing markets for financial
instruments and commodity futures (such as those traded on the Chicago Mercantile Exchange). B2B firms promise to extend such markets to a variety of new
products, including manufactured goods, primary inputs, and services. Exchanges
provide a host of services, including rules for trading, price transparency, and
centralized clearing. Trading rules can be structured so that buyers and sellers
expect to receive the best available price for their transactions. Typically, buyers
and sellers can observe the prices of transactions as they occur. Centralized clearing
reduces transaction costs because buyers and sellers need only settle with the
exchange based on their net position at the end of the day, rather than settling
each transaction individually. B2B exchanges typically deal in unregulated forward
contracts and thus differ from financial futures exchanges, which are governed by
the Commodity Futures Trading Commission and other government agencies.
Examples of B2B companies organized as exchanges are AlmondEx (almonds), Altra Energy (oil and gas), Arbinet (telecommunications bandwidth),
CheMatch (chemicals), e-Steel (steel), and PaperExchange (paper and pulp).
Cantor Fitzgerald, the parent company of eSpeed, provides an important example
60 Journal of Economic Perspectives
of a company establishing an on-line service that will cannibalize its existing
business. Cantor Fitzgerald operates about 50 percent of the global wholesale
market for fixed-income securities such as Treasuries, corporate bonds, and municipal bonds.5
Centralized clearing is a straightforward service in stock exchanges or with
limited numbers of well-defined commodities, but it presents more of a challenge
in B2B markets with a proliferation of different specialized products. With many
different product specifications possible—say, thousands of types of plastic
moldings—exchanges may often look more like small-party negotiations with only
two or three participants, rather than like financial double-auction markets with
thousands of participants trading shares of the same stock in a given day.
In some of the on-line B2B markets we have observed, the exchange has been
organized like a bulletin board: a buyer may post a bid for a desired commodity,
say four tons of grade-3 low-density polyethylene to be delivered to St. Louis
on October 1, with some bid price in dollars. In response, a seller can post a
counteroffer—and rather than merely posting an ask price somewhat higher than
the bid on this commodity, the seller may also decide to change the product
specifications. For example, the seller might post the information that while it has
no grade-3 product available for October 1, it might instead provide a grade-2
product on October 10, at a specified ask price. The original buyer—or a new
buyer—could then respond with a new bid on the newly defined commodity.
Perhaps as the market grows and the number of participants increases, there
will be enough transaction volume to support a separate double auction for each
separate commodity, just as in markets for precious metals or for shares of stock.
Another possibility is that intelligently designed computer software will make it less
cumbersome to conduct these types of negotiations with lots of permutations of
product attributes. Creating an exchange for such products will likely entail specialized auction procedures that adjust prices for multidimensional product attributes. The exchanges also will improve economic efficiency of input markets by
creating standardized products, thus allowing competitive bidding by multiple
buyers and sellers, as in commodity futures markets.
Some firms fall into more than one of these four categories. For example,
PlasticsNet runs auctions for some transactions, but also operates as a broker by
allowing its users to place classified ads for products. Similarly, MetalSite runs
single-sided auctions for some clients, but operates a double-sided exchange as well.
5 Cantor Fitzgerald estimates the total value of fixed income securities around the world at $45 trillion,
(advertisement in the New York Times, May 11, 2000, p. C3). The company offers an electronic trading
alternative to established bond trading, most of which has traditionally been carried out by telephone
conversations (Raynovich, 2000). ESpeed does not plan to disintermediate brokers but rather to offer
them its bond trading technology and service. ESpeed characterizes itself as a business-to-business
market maker, with a service that allows investors to trade bonds and other financial assets just as they
trade stocks (Casey, 2000). By mid-2000, Espeed’s systems handled over $150 billion in daily transactions, well over four times the daily volume of the New York Stock Exchange (see ^http://www.eSpeed.
com&).
David Lucking-Reiley and Daniel F. Spulber 61
Market Structure and Ownership of B2B Intermediaries
At the formative stage of B2B e-commerce, segments of the intermediary
marketplace appeared to be highly competitive. There were hundreds of entrants
with projections of thousands more (Latham, 2000, p. 3). Rapid initial entry
suggests that entry costs were low relative to expected returns. Entry costs also
appeared to be low because companies could rent communications and computer
facilities without incurring irreversible capital costs. Moreover, market entrants
could outsource operation of their website to specialized service providers, the
so-called e-commerce platforms. For example, Ariba and CommerceOne developed software platforms for running e-commerce marketplaces (either via fixedprice catalogs or via bidding to determine prices), and they provided the software
to a number of companies. A wide variety of software applications became available.
Thus, for many B2B companies entry costs were primarily focused on the design of
e-commerce services and on marketing and sales expenditures to attract buyers and
sellers.
However, returns to scale and the importance of liquidity suggest that eventually only one or two markets will operate in each product or service category.
Economies of scale result from the fact that creating an Internet-based market
involves large fixed costs, while the marginal costs of providing transaction information to market participants appear to be near zero. Moreover, as the number of
participants at a site increases, buyers and sellers both find it easier to realize
transactions in a market, so that a greater number of sellers attracts more buyers
and conversely a greater number of buyers attracts more sellers. Accordingly,
buyers and sellers have an incentive to trade on the highest-volume exchange.
These scale and liquidity effects would be reduced if industries managed to settle
on open standards for the exchange of information about products, because then
a smaller exchange could easily gather information and trading partners from a
larger exchange, but it is unclear whether such interoperability between markets
will materialize.
The history of commodity futures markets demonstrates that new markets
often fail. Carlton (1984, p. 256) observes that between 1921 and 1983, 180
different futures contract markets existed, with an average lifetime of less than 12
years. Those contract markets founded after 1921 had an average lifetime of about
nine years, with only three specific commodity contracts continuously listed in the
Wall Street Journal for the entire period studied. An individual futures exchange can
encompass many futures contract markets, so the failure of an individual market
need not alter the lifespan of a futures exchange. Carlton also points out that
futures exchanges compete with each other for volume and the industry has tended
to converge to highly concentrated market structures.
The rate at which concentration takes place in B2B e-commerce will depend
on the speed at which buying and selling migrates to the Internet, what sorts of
intercompany transaction mechanisms prove popular, and what methods evolve for
standardizing transactions of goods that come in many closely related varieties.
62 Journal of Economic Perspectives
Some concentration appears to be occurring globally in equities exchanges. For
example, the Paris, Brussels and Amsterdam exchanges merged to form the
Euronext exchange. This concentration of equities exchanges occurred partly in
response to the efficiency of electronic communications networks such as Instinet,
Island and Archipelago, which had captured 30 percent of Nasdaq trades in the
United States (Morgenson, 1999).
Consolidations in financial markets suggest that the proliferation of B2B
exchanges is likely to be a short-run phenomenon, with eventual concentration of
volume in specific markets through mergers and the exit of smaller exchanges.
Industry reports and our own observations of dozens of e-commerce sites indicate
that less than 15 percent of B2B exchanges were actively operating markets.6 Many
were “vaporware,” premature announcements designed to stake out market territory for companies before their service was actually available. Some observers
predicted substantial shakeouts within several years as the markets become established (King, 2000).
There are several different forms of ownership of B2B e-commerce companies:
independent firms that operate a website, traditional dealers who also operate
on-line markets, and industry-operated exchanges. It is not clear whether one type
of ownership structure will be more successful than others.
Many early B2B e-commerce companies were independent startups financed
by venture capital. Some, like VerticalNet, became publicly traded companies. By
mid-2000, there were at least 600 on-line trading exchanges funded by venturecapital firms (Latham, 2000, p. 3). ChemConnect also included customers BASF
and Dow as equity partners, while PaperExchange included customers International Paper and Staples.
Many established distributors shifted part of their business on-line or took equity
shares of on-line markets, as noted previously. At least to date, there are no prominent
examples of an existing distributor being displaced by a purely on-line business
(Eriksen, 2000, p. 7). DoveBid, a decades-old industrial machinery auctioneer that
entered into on-line auctions, has already established business arms for value-added
transportation, appraisal, and inspection services. The on-line chemical marketplace
Chempoint is a subsidiary of the traditional distributor Royal Vopak. Other large
established dealers entering B2B e-commerce include electronics and computer industry companies such as Arrow Electronics, Avnet, Ingram Micro, and Tech Data.
Alliances of established manufacturers announced plans to sponsor marketplaces. These industry-sponsored exchanges base their strategy on the idea that the
technology of Internet marketplaces is relatively easy to reproduce, and that the
most important asset of an intermediary is the business of its key buyers and sellers.
For example, Weirton Steel, LTV, and Steel Dynamics formed MetalSite, which
conducts thousands of transactions monthly. Sears, Carrefour, and several other
major retailers started GlobalNetExchange to organize purchases from over 50,000
6 Comprehensive, up-to-date lists of B2B e-commerce companies can be found at these helpful reference
sites: ^http://www.line56.com/b2bprofiles/verticals.cfm& and ^http://www.nmm.com/knowledgebase&.
Business-to-Business Electronic Commerce 63
vendors. Boeing, Lockheed Martin, BAE Systems, and Raytheon formed an exchange for aerospace parts and services with the potential for over $70 billion in
business with 37,000 suppliers. Similarly, International Paper (despite its previous
equity investment in PaperExchange) joined with Georgia Pacific and Weyerhaeuser to announce the ForestExchange trading exchange for paper and forest
products. By mid-2000, it was estimated that 60 buyer-dominated consortia, representing over 278 companies and $3 trillion in annual purchasing, planned to
establish their own electronic markets rather than relying on independent exchanges (Roberti, 2000).
The establishment of B2B exchanges has raised several antitrust issues. Antitrust
policymakers are concerned that B2B websites will allow competitors to exchange price
information, thus facilitating collusion to fix prices. Also, antitrust authorities worry
that B2B websites will foreclose participation by competitors, leading to their exclusion
from segments of the market. The FTC negotiated with the automakers regarding the
independence of the industry-sponsored parts exchange Covisint (Leonhardt, 2000).
These concerns recall antitrust scrutiny of airline computer reservation systems that
alleged collusion, exclusion and biased listings. Buyer- or seller-dominated websites
could become mechanisms for the exercise of market power.
In industry-sponsored exchanges as they have been announced, ownership
tends to be on the side of the market with the greatest concentration of market
power; for example, few buyers and many suppliers in automotive parts, many
buyers and few suppliers in paper products. In contrast, smaller buyers and sellers
may favor independently-owned exchanges offering transparency in execution of
trades and up-to-date pricing information (Brown, 2000). Competition between
exchanges should create incentives to avoid foreclosure or the exercise of market
power. Owners of exchanges have incentives to increase volume and to foster
liquidity by attracting more buyers and sellers, which should in turn lead to the
development of independent neutral exchanges.7
Effects of E-Commerce on the Organization of Firms
Ronald H. Coase’s (1937) classic article introduced the concept of transaction
costs. Coase explained that the costs of using the market were an important
determinant of whether firms would carry out an economic activity within their
organizations or rely on purchases from other firms. When using the market is
costly relative to management costs, companies have an incentive to vertically
integrate. Yet outsourcing is compelled by the buyer’s need for flexibility and focus,
supplier economies of scale and scope, and supplier expertise. To the extent that
7 According to AMR Research (2000), Covisint “has come to the understanding that its main customers
aren’t GM, Ford or DaimlerChrysler, but the suppliers.” According to AMR Research, Consortium
Trading Exchanges “must behave like vendors and actively market themselves, manage expectations and
deliver value in order to survive.”
64 Journal of Economic Perspectives
e-commerce technology lowers the costs of intercompany transactions, it should tip
the balance toward greater use of external markets.
The potential effects of B2B e-commerce extend beyond saving money on
transactions between existing firms. Cost and allocative efficiencies in e-commerce
suggest a more fundamental change in the way that businesses are organized.
Vertically integrated firms engage in substantial internal sales and procurement
activities. With B2B e-commerce, such vertically integrated companies might reorganize to outsource production of goods and services that were previously produced internally, as well as outsourcing the management of these transactions. For
companies that purchase externally, B2B e-commerce intermediaries would substitute for many of the activities of company purchasing, sales, marketing, and even
accounting departments. As market transaction costs fall with the maturation of
business-to-business e-commerce, outsourcing and vertical disintegration will occur,
resulting in more independent entities along the supply chain.
The automobile industry offers a vivid example of this change. At the beginning
of the twentieth century, the automobile industry exhibited a strong preference for
vertical integration. Ford had a slogan: “From Mine to Finished Car, One Organization” (Casadesus-Masanell and Spulber, 2000). By 1920, General Motors “had extended
its scope so that not only all the engines used in its cars, but a large proportion of such
units as gears, axles, crankshafts, radiators, electrical equipment, roller bearings, warning signals, spark plugs, bodies, plate glass, and body hardware, were produced either
by a General Motors unit or by a subsidiary” (Edmonds, 1923).
But by the beginning of the twenty-first century, the automobile industry had
begun to reevaluate its organizational structure. GM had spun off its parts manufacturing unit Delphi Automotive Systems Corp., thus creating the world’s largest
auto parts supplier with over 200,000 employees. Similarly, Ford had spun off its
Visteon parts division in summer 2000, converting internal transactions into B2B
transactions. GM, Ford, and DaimlerChrysler established a company called Covisint
to handle auto parts transactions from suppliers. The supply chains of these three
companies total almost $250 billion, so Covisint could become one of the largest
businesses on the Internet (Moore and Trenker, 2000). The pattern of greater
reliance on coordination through markets and less emphasis on vertical integration
and organizational governance seems likely to proliferate.
Conclusion
Advances in computers and communications clearly hold great promise for
reducing transaction costs between businesses. Productivity gains may result from
the automation of transactions, the potential economic advantages of intermediation, the organization of centralized exchanges, and the reorganization of firms.
An important research question is the measurement of these economic efficiency gains. Estimation of productivity growth in services such as B2B e-commerce
David Lucking-Reiley and Daniel F. Spulber 65
presents difficulties.8 Triplett and Bosworth (2000) observe that economic changes
attributable to e-commerce cross the traditional production boundary used in
national accounts. As an example, they compare the purchase of a book from a
traditional retailer with the purchase of a book from an on-line retailer. Comparing
the prices in the two settings ignores the costs of travel and time involved in visiting
the traditional retailer, while explicitly counting the costs of shipping and handling
for the on-line purchase. In the case of B2B e-commerce, one would like measurement of productivity to reflect total net benefits, including lower search and
procurement costs for buyers and sellers. A related question is how much of the
current activities of companies in manufacturing, construction, energy, transportation is devoted to production operations and how much is attributable to transactions.
One of the challenges faced by companies in B2B e-commerce is the development of software and communications standards. Extensible Markup Language
(XML) is being applied to develop data descriptions and protocols to describe
practically all aspects of a transaction, including product features, transportation,
prices, and credit terms. If standards are widely adopted, manufacturers, suppliers
and distributors will be able to exchange commercial information using generally
recognized formats (Mitchell, 1999; Bosak and Bray, 1999). Such standardization
enables the computers of both parties to a transaction to understand precisely what
is being traded, so that each party can automatically update its internal records,
such as billing and inventory. Developing such protocols will require extensive
cooperation of buyers and sellers within industries.9 An important question for
economic observers is what types of goods and services can be standardized and
what types of transaction protocols will emerge.
Business-to-business e-commerce seems likely to engender a significant reorganization of industry, with a certain amount of vertical disintegration and new
roles for intermediaries and market makers. The extent to which e-commerce will
change the organization of firms has empirical significance for the theory of the
firm. Another important issue is what types of market mechanisms will be favored
by e-commerce and the relative importance of different types of intermediaries.
Will B2B exchanges be owned by industry consortia or will they tend toward
independence?
The plethora of entrants and business models, and the significant returns to
market consolidation, suggest that substantial entry and exit of firms will take place
before the benefits of B2B e-commerce are obtained. The economic significance of
intercompany transactions suggest that even small enhancements in the efficiency
of transactions will eventually produce extremely large cost savings in the economy.
8 For a discussion of the overall effect on productivity of advances in computers and communications,
see the “Symposium on Computers and Productivity” in the Fall 2000 issue of this journal.
9 RosettaNet is an example of an organization devoted to the coordinated development of industry
standards for information interchange.
66 Journal of Economic Perspectives
y David Lucking-Reiley gratefully acknowledges the support of the National Science Foundation through grant SBR-9811273 and the hospitality of the Kellogg Graduate School of
Management. Daniel F. Spulber gratefully acknowledges the support of the Searle Fund. We
thank Zeev Klein and Carey Rountree for research assistance. We particularly thank Timothy
Taylor, Brad De Long, Alan Krueger and Michael Waldman for useful editorial suggestions.
We also thank Elise Ackerman, Charles Allen, Don Bielinski, Maggie Bray, Doug Bryan, Luke
Froeb, Shane Greenstein, Austan Goolsbee, Tom Gros, Jay Hawthorn, Mark Kiefer, Toma´s
Junqueira de Camargo, Julie Landry, Lewis Pennock, Liron Petrushka, Laura Pilz, Naghi
Prasad, Kathryn Shaw, and Gregg Speicher for helpful conversations. We have benefitted from
seminar presentations at the Harvard Business School Conference on Competition and
Strategy and the Decentralization Conference at IBM