[Economic Report of the President (1998)]
[Administration of William J. Clinton]
[Online through the Government Printing Office, www.gpo.gov]

[DOCID: f:erp_c6._]
Economic Report of the President - - - - - - - - - - - - H. Doc. 105-176
[From the online service of the the U.S. Government Printing Office]
[wais.access.gpo.gov]



CHAPTER 6 -- Recent Initiatives in Antitrust Enforcement

DURING THIS ADMINISTRATION the Federal antitrust enforcement agencies
have been aggressive in enforcing the Nation's antitrust laws. The
Antitrust Division of the Department of Justice has imposed record
fines-over $200 million in fiscal 1997-and the Justice Department and
the Federal Trade Commission (FTC) have both pursued many important
cases and investigations, involving such firms as Microsoft, Archer
Daniels Midland, Toys ``R'' Us, and Staples and Office Depot, as well as
traders on the NASDAQ over-the-counter stock market. This more
aggressive stance does not, however, return Federal antitrust
philosophy to an earlier era in which big was viewed as inherently
bad. Recent cases and investigations suggest that the Justice
Department and the FTC have taken a balanced approach to antitrust
enforcement, bringing an action only when thorough investigation and
analysis reveal a substantial threat to competition. In doing so,
these agencies are guided by their mission to protect the competitive
process, recognizing that free markets are likely to provide the best
outcomes for society.

This chapter reviews how these agencies have analyzed market
competition in a number of recent cases. In so doing it attempts to
explain some apparent paradoxes in antitrust enforcement-why, for
example, in 1997 the FTC stopped Staples and Office Depot from
merging, even though the vast majority of office products are sold by
neither company, but allowed a merger between the two leading U.S.
manufacturers of large commercial aircraft in an already highly
concentrated industry. The chapter begins with a broad overview of the
origins and principles of antitrust efforts in the United States and
then proceeds to survey several recent developments. The most striking
of these has been the growth in corporate merger filings to record
levels. The chapter explores the efforts of the antitrust enforcement
agencies to allow those mergers that reduce costs, without allowing
firms to gain the power to raise prices. Next the chapter discusses
the potential impact of electronic commerce on competition. Although
electronic commerce will in many cases make competition work more
smoothly, it may also make it easier to establish price-fixing
agreements. The chapter also surveys the efforts of antitrust
enforcers to ensure the continued growth and competitiveness of
high-technology industries. Finally, the chapter discusses
international antitrust enforcement, an aspect of antitrust policy
that has become increasingly important as global trade has expanded.

ORIGINS AND PRINCIPLES OF ANTITRUST

As the American economy shifted from agriculture toward industry
during the 19th century, large corporations and trusts began to
emerge, eventually dominating or threatening to dominate a number of
industries. Public opposition to these monopolies mounted, and in 1889
alone, 12 States passed antitrust or antimonopoly statutes. The
Congress followed swiftly. In 1890 it passed the Sherman Act by an
overwhelming margin: 52 to 1 in the Senate and 242 to 0 in the House
of Representatives. The broad contours of American antitrust law were
completed in 1914 with the passage of the Federal Trade Commission Act
and the Clayton Act.

The Sherman Act contains broad bans-with both criminal and civil
penalties-on monopolization, price-fixing agreements, and other
unreasonable restraints on trade. The Clayton Act contains more
specific prohibitions of mergers and of certain forms of price
discrimination, exclusive dealing agreements, and tie-in sales (sales
conditioned on the purchase of another product) when the effect may be
to substantially lessen competition or to tend to create a monopoly.
The Justice Department and the FTC have overlapping but distinct
authorities: the Justice Department may bring actions under the
Sherman Act and the FTC under the Federal Trade Commission Act, but
either may bring actions under the Clayton Act. In addition, the major
regulatory agencies, such as the Federal Communications Commission,
the Federal Energy Regulatory Commission, and the Surface
Transportation Board, all review mergers under their own statutory
authority.

The antitrust laws' primary objection to monopolies, cartels, and
other restrictive practices and restraints of trade is that they
injure consumers by increasing prices. Another concern, which has been
a particular focus of economists, is that these high prices
inappropriately curtail consumption of the monopolized good.
Inefficiencies arise when sellers charge monopoly prices, because
consumers lose more from the price increase than sellers gain.

Another objection to monopoly was expressed by Judge Learned Hand, who
argued that ``Unchallenged economic power deadens initiative,
discourages thrift and depresses energy,'' and that ``immunity from
competition is a narcotic, and rivalry is a stimulant, to industrial
progress.'' In a similar vein, the British economist John Hicks wrote
that ``the best of all monopoly profits is a quiet life.'' This
complacency on the part of monopolists can impede economic progress.
The concern that firms with market power-the power to raise prices
above their production costs-can limit innovation has become an
important part of antitrust enforcement during this Administration.

The choice between competition and monopoly is easy.
Unfortunately, however, that is not usually the choice that antitrust
enforcers face. The industries in which antitrust issues tend to arise
can seldom be appropriately classified as either perfectly competitive
or monopolized. Usually they lie somewhere in between. Firms typically
have some market power, but they also have competitors. Mergers and
restrictive practices may create or enhance market power, but they may
also promote efficiencies and hence can benefit consumers. Identifying
corporate conduct whose primary effect is to lessen competition is the
task of antitrust enforcers-a task that often presents a formidable
analytical challenge.

MERGERS

Another challenge for the antitrust enforcement agencies during this
Administration has been the dramatic increase in merger activity. As
Chart 6-1 shows, after a lull in the early 1990s the merger market has
come roaring back to life. Both the 1996 and 1997 fiscal years set new
records for the number of merger filings.



-------------------------------------------------------------------------

Box 6-1.-Consolidation in the Defense Industry

The recent merger wave in the U.S. defense industry highlights the
difficult tradeoffs involved in antitrust policy and the balanced
approach that the antitrust enforcement agencies have taken during
this Administration. The end of the Cold War and the ensuing
65-percent real reduction in the Pentagon's procurement budget created
intense pressure toward consolidation. A large share of the defense
business is now concentrated in the hands of a few large firms-notably
Lockheed Martin, Northrop Grumman, Boeing, Raytheon, and General
Dynamics-that have acquired numerous other major defense contractors
as they exited the industry.

The challenge for the antitrust authorities has been to balance the
perceived need for consolidation to reduce overhead costs against the
potential for a reduction in competition. On the one hand, if the
mergers allow defense contractors to eliminate duplicative overhead
costs the Pentagon will be able to purchase weapons systems more
cheaply. On the other hand, if the number of effective bidders falls,
prices may rise, forcing either higher defense budgets or reduced
defense purchases.

In a number of cases where anticompetitive effects have been a
concern, instead of trying to block the merger and forgoing the
potential cost savings, the antitrust agencies have tried to adopt
narrowly focused remedies. For example, they have invalidated
exclusivity arrangements, insisted on the divestiture of key assets,
and required the creation of provisional information
-------------------------------------------------------------------------

In evaluating these mergers and deciding which ones to challenge, the
enforcement agencies must strike a fine balance. A merger may yield
significant cost savings, but it may also threaten to increase
industry concentration (that is, reduce the number of firms in the
industry) and stifle competition, allowing the remaining firms to
increase prices and reduce output. The impact on concentration and
competition is particularly difficult to evaluate in the many
industries now experiencing rapid structural and technological change,
such as the defense industry, considered in Box 6-1. The enforcement
agencies must consider who will be the merged firm's competitors in
the future, not just today.
A merger does not have to create a monopoly in order to result in
higher prices and lower output. By increasing concentration, a merger
may increase the likelihood of successful collusion, either overt or
tacit, among the remaining firms. Greater concentration may make it
easier for each firm to communicate its intentions to the others, and

-------------------------------------------------------------------------
Box 6-1.-continued
``firewalls'' between merging companies. An example of the first remedy
is provided by the 1995 merger of Lockheed Corp. with Martin Marietta
Corp. This merger raised antitrust concerns because the companies had
entered into exclusive teaming agreements with Hughes and Northrop
Grumman Corp., respectively. In the wake of a merger, these agreements
would raise the prospect that there might be only one bidder on
space-based infrared early warning satellite systems, since Hughes and
Northrop Grumman were the leading providers of electro-optical sensors
for these satellites. To promote competition in this market, the FTC's
consent order forbade Lockheed Martin from enforcing the exclusivity
provisions.

Likewise, Raytheon Co.'s $5.1 billion acquisition of Hughes
Aircraft Co. might have substantially lessened competition in both
infrared sensors and electro-optical systems had the Justice
Department not forced Raytheon to make a large divestiture. Raytheon
agreed to sell off the infrared sensor business it had acquired from
Texas Instruments Inc., as well as electro-optical systems businesses
that it would otherwise have acquired with the purchase of Hughes.
Raytheon also agreed to a firm price on an Air Force missile to
compensate for the lost competition from Hughes. Finally, Raytheon
agreed to maintain an information firewall to preserve the
independence of Raytheon and Hughes as competitors for a new Army
antitank missile.
-------------------------------------------------------------------------

the interests of the firms may be less likely to diverge. The smaller
number of firms may also reduce the benefits and increase the cost of
cheating on the collusive agreement. For example, mergers make price
cutting less profitable because the merger eliminates one firm from
which customers might be attracted away by the price cut. It may also
become easier for colluding firms to detect and punish those firms
that deviate from the agreement.

Mergers may result in price increases even when firms do not collude
in any sense. For example, a firm with market power by virtue of
control over a large portion of industry capacity will enhance that
power, and may therefore raise prices, if it acquires still more
capacity by merging with a competitor. Another important example of
such a ``unilateral competitive effect'' arises when formerly
standardized products become differentiated, giving rise to market
power as consumers develop brand preferences. Such power is limited by
the availability of competing brands; hence a merger between firms
selling competing brands relaxes the constraint that competition
places on prices. The merged firm recognizes that some of the sales
lost through a price increase on one brand will be recaptured by the
other brand, and therefore be retained by the firm. This encourages
the merged firm to raise the price of both brands. When the brands are
particularly close substitutes, the firm may want to raise both prices
substantially.

Enforcement agencies must balance these concerns about market power
against the efficiencies in production that mergers can make possible.
There are several ways in which mergers can reduce the average cost of
production in an industry. A merger may allow one firm to take
advantage of another's superior technology. Where production processes
are composed of multiple distinct activities, a merger can allow each
of the merging firms to specialize in those activities that it does
best. Mergers may also increase efficiency in industries subject to
economies of scale, that is, those in which average production cost
declines as output increases. In these industries a merger may reduce
costs by eliminating duplicative fixed costs or allowing longer
production runs.

Consumers benefit from the merger as well if merging firms pass these
savings along in the form of lower prices. The challenge for antitrust
enforcement, then, is to prevent those mergers that would harm
consumers by enhancing market power, but to allow those that create
substantial benefits. To evaluate the market power and the efficiency
effects of mergers, the FTC and the Justice Department use the
framework that they jointly established in the 1992 Horizontal Merger
Guidelines, which were partially revised in 1997. According to the
guidelines, the steps to be taken in a merger review for a merger
among competitors are as follows:

 define the relevant market and calculate its concentration before
and after the merger

 assess whether the merger raises concerns about adverse
competitive effects

 determine whether entry by other firms into the market would
counteract those effects, and

 consider any expected efficiency gains.

This chapter discusses each step in turn below.

MARKET DEFINITION

The first step is to determine the relevant market and whether the
merger will increase concentration significantly in that market. The
Merger Guidelines state that the relevant market is generally the
smallest group of products and geographical area such that a
hypothetical monopolist in that market would raise the price
significantly, taking into account the reduction in demand caused by
consumers curtailing their purchases. Having defined the relevant
market, the agencies determine the market shares of all firms
identified as market participants, and use these market shares to
calculate an index of market concentration. Mergers that would
increase concentration significantly tend to attract more scrutiny
from the enforcement agencies, because these mergers are more apt to
lead to large price hikes. Typically, therefore, the narrower the
relevant market, the more likely it is that a merger will be
investigated.

In 1997 the FTC challenged the merger of Staples Inc. and Office Depot
Inc. because it believed that the relevant product market was ``the
sale of consumable office supplies through office superstores,'' and
these firms were the two largest in that market. Staples countered
that the relevant market was all sales of office products, including
sales by discount stores, drugstores, and wholesale clubs. The
combined firm would have accounted for less than 6 percent of this
broader market, which suggested that the firm could not have raised
prices significantly after the merger if this market definition were
indeed correct. The FTC maintained, however, that even though most
individual items could themselves be bought from many retailers, the
size, selection, and inventory offered by office superstores
distinguish them from other office supply retailers. The FTC's
statistical analysis showed that, when the presence of other potential
competitors was controlled for, Staples' prices were over 5 percent
higher in cities where it did not face competition from other office
supply superstores. The FTC took this as evidence that nonsuperstore
sellers of office supplies do not constrain superstores' prices
effectively. This pricing evidence led the court to accept the FTC's
market definition and conclude that the merger would significantly
increase concentration in the office superstore market and so be
anticompetitive.

The key issue in defining the relevant market in a recent merger
between two gypsum drywall producers was not the type of seller, but
rather the sellers' geographic location. In 1995 Georgia-Pacific
Corp., which had 10 drywall plants nationwide, including one each in
New York and Delaware, proposed to acquire nine drywall plants from a
Canadian-based competitor, Domtar Inc. Two of the nine plants were
located in New Hampshire and New Jersey. The Justice Department
determined that if the relevant geographic market was national, the
acquisition would likely not have raised competitive concerns.
However, the merger would have increased concentration significantly
in the Northeastern States, so that if the relevant market were
localized to that region, the merger likely would have led to price
increases there.

To determine the relevant geographic market, the Justice Department
examined whether a small but significant local price increase by
Northeastern producers would be profitable, taking into account the
extent to which customers could switch to producers outside the
region. The agency considered such factors as current shipping
patterns, constraints on production capacity outside the region, and
transportation costs. Gypsum drywall is heavy, bulky, expensive to
ship, and likely to break during transport if handled excessively. The
Justice Department found that drywall plants in the Northeastern
States accounted for the majority of sales to consumers in those
states; sales from plants outside the region were comparatively small.
Furthermore, drywall plants outside the Northeast had relatively
little excess capacity. From this evidence, the Justice Department
determined that customers in the Northeast could not have switched to
out-of-region producers in sufficient quantities to make a local price
increase unprofitable. The agency therefore decided that the relevant
geographic market was regional, and Georgia-Pacific, to satisfy the
Justice Department's concerns, agreed that it would divest its New
York and Delaware plants.

COMPETITIVE EFFECTS

Defining the market and assessing its concentration are only the
beginning of the merger review process. The next step is to determine
whether the merger would have adverse competitive effects. The 1992
Merger Guidelines recognize that mergers may lessen competition
through either collusion or unilateral effects.  Indeed, unilateral
effects received new prominence in the 1992 Merger Guidelines and have
been the dominant concern in several recent mergers.

One recent example where the analysis of unilateral effects suggested
significant harm to competition is the acquisition of Continental
Baking Co. by Interstate Bakeries Corp. Continental's Wonder Bread
brand competed against various Interstate brands in several regions.
Although these two firms were by no means the only producers of white
bread in these regions, the Justice Department concluded that white
bread is a highly differentiated product, with various brands
commanding significant customer loyalty, and that after the merger
Interstate would likely have raised prices on its brands even if other
bakers kept their prices constant. Interstate would no longer be
discouraged from raising prices on its own brands by the risk of
customers switching to Wonder Bread, since after the merger Interstate
would own Wonder Bread. Likewise, whereas Continental was discouraged
from raising the price of Wonder Bread by the prospect of customers
switching to Interstate's brands, after Interstate bought Wonder Bread
this would no longer be a worry. Simulations based on estimated demand
elasticities helped convince the Justice Department that significant
price increases would likely follow the merger, even in the absence of
coordination among the remaining firms. To avoid these price
increases, the Justice Department entered into a consent decree
requiring the merged firm to divest a brand of bread in each of five
geographic regions.

Sometimes the antitrust authorities can limit a merged firm's power to
raise prices without requiring a divestiture, as illustrated in the
merger of Time Warner Inc. with Turner Broadcasting System Inc. In
1995 Time Warner proposed to acquire Turner in a deal valued at over
$7 billion. Both companies were important providers of programming to
local cable system operators. Time Warner owned Home Box Office (HBO),
the leading cable movie channel, and Turner owned Cable News Network
(CNN). Both these channels are ``marquee'' channels that cable operators
have a strong desire to carry in order to attract and retain
subscribers. The FTC was concerned that if Time Warner controlled both
these marquee channels it would increase the prices it charged to
cable operators. To limit the anticompetitive effects of the merger,
the FTC's consent order prohibited Time Warner from ``bundling'' HBO
with Turner channels, and CNN with Time Warner channels. The bundling
restriction required that the Time Warner and Turner channels be
offered separately at prices that do not depend on whether the other
is purchased.

It may not be immediately apparent why restrictions on bundling can
sometimes be an appropriate remedy; after all, once the merged firm
controls the price of both channels it could simply implement an
across-the-board price increase. However, a hypothetical example
demonstrates that when a merged firm sells goods that are substitutes
for each other, prohibiting bundling can limit price increases.
Consider a cable operator in a city with 50,000 potential subscribers,
and assume that the cable operator earns a dollar in profits from each
subscriber. Suppose that 20,000 of the potential subscribers like
movies: they will subscribe only if the cable system offers a movie
channel. Another 20,000 like news and will subscribe only if a news
channel is offered. The remaining 10,000 like both movies and news and
will subscribe if either is offered. In this city the cable operator
would be willing to pay up to $30,000 for either movies or news, since
in each case 30,000 people will subscribe. However, as soon as the
cable operator buys a movie channel and gets all the subscribers who
like movies, it will be willing to pay only $20,000 for a news
channel, since the only additional subscribers it will attract are the
20,000 people who like news but not movies. Similarly, a cable
operator that already offers news would be willing to pay only $20,000
for movies. Since some people subscribe if either a movie channel or a
news channel is offered, the two channels are substitutes from the
point of view of the cable operator. If movies and news can be sold as
a bundle, they can be sold for $50,000, because a total of 50,000
people will subscribe. On the other hand, if bundling is forbidden and
each channel must be for sale individually, the merged firm will not
be able to charge that much.

Suppose, for instance, that the merged firm tried to sell each channel
for $25,000. The cable station would respond by buying only one of the
channels; since the channels are substitutes, once the cable station
purchases one channel, its willingness to pay for the other channel
diminishes to $20,000. If the channels are sold separately, the most
the merged firm could sell them for is $40,000 ($20,000 each). For
this reason, restrictions on bundling such as those in the FTC's
consent order can sometimes limit the exploitation of market power,
even when the firm can charge whatever it likes for its products
individually.

ENTRY

The analysis of a merger does not end with defining the market and
determining whether the increase in concentration would allow the
merged firm to raise prices. Entry can in principle constrain the
merged firm's ability to raise prices: a merger that leads to
increased prices may also create opportunities for new firms to enter
the market, charge a lower price to gain market share, and still earn
profits. Loss of sales to new entrants could cause the anticompetitive
price increase to be unprofitable. As a result, entry or the threat of
entry can in some cases prevent any appreciable price increase after a
merger.

One difficulty with entry analysis is that it can be highly
speculative. It is easy to be overly optimistic and assume that
entrants will materialize and eradicate the anticompetitive effects of
a merger. Accordingly, the antitrust enforcement agencies have taken
seriously the Merger Guidelines' caution that entry must be timely,
likely, and sufficient to counter the merger's adverse competitive
effects.

One merger where entry seemed unlikely to offset the effects of
increased concentration was the proposed 1995 acquisition of Intuit
Inc. by Microsoft Corp. Each of the two software firms produced a
popular personal finance program: Microsoft's Money and Intuit's
Quicken together accounted for more than 90 percent of the personal
finance software market. Here the question faced by the Justice
Department was whether other firms were likely to enter this market in
sufficient force to constrain Microsoft's market power once it owned
both programs. Two important features of software markets limited the
likelihood of entry: the importance of reputation and the ``lock-in
effect.'' Purchasers of personal finance software generally prefer a
product that is widely accepted as reliable and successful and that
has a reputation for performance and customer support. It can take
many years and a significant investment for an entrant to develop such
a reputation. Even Microsoft had considerable difficulty overcoming
the initial success of Intuit. After 4 years of effort, the market
share of Microsoft's Money remained far less than that of Quicken, and
Microsoft had yet to achieve a positive return on its investment. The
fact that consumers have to put considerable time and effort into
learning to use a given program gives rise to the lock-in effect.
Users of existing software may be reluctant to incur the switching
costs of learning another program. Future purchasers may likewise
hesitate to invest time and effort in learning to use an entrant's new
and untested product because of the risk that the product may not
succeed in the marketplace, requiring the customer to eventually
switch to the established product.

To make the deal acceptable to the antitrust authorities, Microsoft
planned to transfer part of its assets in Money to another software
developer. Even so, the Justice Department felt that the importance of
reputation and the lock-in effect, among other factors, meant that
entry could not be relied upon to offset the high concentration that a
merger of Microsoft and Intuit would have caused. The merger was
challenged, and Microsoft decided not to pursue it.

EFFICIENCIES

The final major step in the merger review process is to consider the
efficiencies promised by the merger. Economists have long recognized
the potential benefits of such efficiencies, and in recent years the
antitrust agencies have been increasingly willing to consider these
benefits when reviewing mergers. Most recently, in April 1997 the
Justice Department and the FTC issued revisions to the section of the
Merger Guidelines devoted to efficiencies. These revisions reflect the
balanced approach of current antitrust enforcement. Under the revised
guidelines, the agencies consider the creation of efficiencies, but
only verifiable, merger-specific efficiencies. Many studies have
suggested that mergers may not produce the synergies and cost savings
claimed by managers. Since the agencies understand that it is easier
for firms to claim efficiencies than to realize them, they subject
efficiency claims to careful scrutiny. If the agencies determine that
the claimed efficiencies are likely to be realized and are of
sufficient magnitude that the merger is not likely to be
anticompetitive, they will not challenge the merger.

The proposed merger between Staples and Office Depot illustrates the
increased consideration and scrutiny of efficiencies in antitrust
enforcement. The two firms claimed that by merging they would be able
to take advantage of large cost reductions and efficiencies in
purchasing, distribution, operations, and marketing, and that these
savings would be passed on to customers in the form of lower prices.
Consistent with the revised Merger Guidelines, the court deciding the
case considered whether these efficiencies would offset the presumed
anticompetitive effects of the merger. The court refused to accept
cost savings that were not merger-specific and dismissed those that
could not be verified. Also at issue was the degree to which Staples
and Office Depot would pass any cost savings through to consumers. The
companies projected that for every dollar of cost savings their prices
would go down by about 67 cents. However, the FTC presented evidence
that historically Staples had passed through only 15 to 17 percent of
its achieved cost savings. Accordingly, the court found that the
merger's efficiencies would not offset its anticompetitive effects. It
granted the FTC's request for an injunction, leading Staples and
Office Depot to terminate their merger agreement.

ELECTRONIC COMMERCE

The potential impact of electronic commerce on competition is
dramatic, as described in a recent White House report titled A
Framework for Global Electronic Commerce. Electronic commerce is
already common in several industries. Travelers, for example, buy
airline tickets from travel agents who use computer reservation
systems. Over-the-counter stocks are traded on a computerized system.
And consumers can buy everything from books to automobiles over the
Internet.

The potential for electronic commerce to make the economy function
better is clear. Computer networks can inform buyers about products
available in other States or, just as easily, in foreign countries.
Cheap information about wide-ranging markets means that buyers can buy
products that they would not otherwise have known about, and can pay
lower prices as well. A seller who is the only supplier in a given
area may have little power to raise prices if buyers can easily
compare prices around the country or around the world. Music stores in
Philadelphia will find it pointless to conspire to sell compact discs
at high prices if buyers can easily locate competing dealers around
the country. Putting cheap information in the hands of consumers thus
seems likely to make markets more competitive. One might well wonder
if electronic commerce could lessen the need for antitrust enforcement
in many markets.

However, two cases that the Justice Department recently filed and
settled-one against a group of U.S. airlines, and the other against
so-called market makers who execute over-the-counter stock
trades-highlight a straightforward problem with electronic commerce.
Computers do increase the information available in the marketplace,
but not just to consumers; they also make more information available
to producers and other sellers. Sellers may be able to use this wealth
of information to form or maintain cartels.

For a cartel to raise prices successfully, the members must somehow
come to an agreement about what prices to charge and must figure out a
way to maintain that agreement. The airline and stock trading cases
illustrate how computer networks can sometimes help a cartel solve
both these problems. They suggest that, rather than lessening the need
for antitrust authorities, the growth of electronic commerce may in
some cases increase it.

In 1994 the Justice Department reached a final settlement in a
price-fixing case involving eight major airlines and the Airline
Tariff Publishing Company (ATP). According to the Justice Department,
the airlines had used ATP's computerized fare dissemination services
to negotiate increases in fares and to trade fare changes in certain
markets for changes in other markets.

The alleged collusive arrangement worked as follows. Each airline
submitted its fare changes or planned future changes to ATP. In turn,
ATP reported the changes to all the other airlines. The resulting data
base was enormous, as each airline offered numerous fares, under
various terms and conditions, on each of thousands of city pairs.
Moreover, these fares changed frequently. In such a complex system it
would seem difficult for the airlines to negotiate or maintain any
price-fixing agreement, much less a covert one. With so many
interrelated fares and fare changes, one might ask how one airline
would distinguish, for example, whether another's price change was an
attempt to cheat on a collusive agreement, an attempt to punish a
third airline for deviating from an agreement in another market, or
simply a normal response to increased costs. The Justice Department
alleged that such confusion was avoided by linking fare changes with
alphanumeric footnote designators and by the judicious use of first
ticket dates. Since the ATP data were computerized, this mass of
information could be analyzed by sophisticated computer programs each
day. Aided by these computer analyses, airlines could engage in
intricate but camouflaged negotiations and could monitor cheating on
agreements. The settlement that the Justice Department entered into
with the airlines barred them from using footnote designators, first
ticket dates, and other devices to communicate with each other.

According to one study, price leadership in the airline industry cost
air travelers $365 million per year during the 1980s. Others have
estimated that the cost of such behavior in the airline industry, had
it been left unchecked, could have reached several billion dollars per
year. These figures suggest that the Justice Department's attempts to
eliminate anticompetitive practices in the airline industry could
yield large dividends for consumers.

The stock trading case, which resulted in a 1996 consent decree,
involved transactions in over-the-counter stocks over the automated
quotation system operated by the National Association of Securities
Dealers (the NASDAQ system). This case also revealed how computerized
information networks can sometimes make it easier for firms to
maintain agreements to sell at high prices. When an investor places a
buy or sell order for shares of a company traded on NASDAQ, special
traders called ``market makers'' typically execute the trade. These
intermediaries make their profits from the bid-ask spread, the
difference between the price at which they buy a stock and the price
at which they sell it.

In the NASDAQ case the Justice Department alleged that NASDAQ market
makers had agreed to a strategy, or convention, for quoting stocks
that essentially limited their incentives to narrow spreads. Also
working to support the agreement was the fact that the NASDAQ computer
network provided sellers with ready (essentially instantaneous)
information about the strategies other sellers were using to quote
prices. Market makers that were observed to deviate from the
convention were harassed by other market makers and threatened with
economic harm.

Traditional economic theory predicts that the price-fixing agreement
alleged by the Justice Department and the Securities and Exchange
Commission (SEC) could not have been maintained on NASDAQ, because
entry barriers were low and any of over 100 firms could enter the
market for any security. If a price-fixing agreement kept the bid-ask
spread high, some market maker would have been tempted to offer the
security at a price below the best asking price, or to buy it at a
price above the best bid, in an effort to increase market share. But
the rules and common practices that governed the way in which NASDAQ
securities were traded could have combined to deter market makers from
undermining the agreement in this way, with the computer network used
for trades playing a key role.

NASDAQ market makers may decline to trade a security at the price
quoted by other market makers. But if a market maker does execute a
trade, the NASD's best-execution rule requires it to make the trade at
the best price quoted on the NASDAQ network. A key feature of trading
on NASDAQ is the widespread practice of preferencing. A preferencing
arrangement between a broker and a market maker commits the market
maker to execute trades submitted by the broker. In combination with
the NASD's best-execution rule, this practice could have sharply
limited the benefits that any market maker could have anticipated from
cheating on any anticompetitive agreement, and so significantly
enhanced the ability of a cartel to maintain collusion. A market maker
that attempted to cheat on an agreement would not expect to
significantly increase its market share, because other firms would, in
effect, match its prices instantaneously and retain their preferenced
order flow. Thus, a practice that initially seemed to offer a great
deal to investors-a guarantee of the best price available, regardless
of which market maker executes the order-in fact may have tended to
support an anticompetitive agreement.

In 1996 the Justice Department entered into an agreement with NASDAQ
market makers. The market makers agreed not to fix prices in the
future and to commit resources to an ongoing monitoring effort to
ensure that they adhere to the antitrust laws.

The lesson of the airline and NASDAQ cases is that computer networks
can sometimes make it easier for sellers to form and maintain
price-fixing agreements, by providing sellers with information about
the prices that other sellers charge. Agreements negotiated by posting
prices on computer networks may prove difficult for the antitrust
authorities to ferret out. In the airline case there was sufficient
ancillary information-in particular the use of annotations linking one
fare proposal to another-to convince the Justice Department that a
negotiation was taking place. In contrast, when one firm tries to take
advantage of the fact that prices can be quickly and easily changed on
computer systems and raises price for a few instants (at a small cost)
in the hope that others will follow, most antitrust experts believe
that there is no violation of antitrust laws, even if other firms do
follow. Simple price leadership is not banned by the Sherman Act, in
part because there is no adequate way to frame a remedy. Firms in
collusion and firms in competition may both move prices in concert.
Antitrust authorities can only try to prevent sellers from negotiating
and offering each other mutual assurances in order to form
price-fixing agreements.

It has always been difficult to tell whether firms are being forced by
competition to charge the same prices, or whether they have agreed to
fix prices. The task could become steadily more troublesome as the
electronic age progresses. Antitrust authorities in the electronic age
need to maintain vigilance in seeking out and enjoining illegal
agreements. Electronic commerce may make antitrust enforcement more
challenging-and more important.

HIGH-TECHNOLOGY INDUSTRIES, INNOVATION, AND INTELLECTUAL PROPERTY

Many of the fastest growing and fastest changing U.S. industries are
to be found in such high-technology fields as aerospace, computer
hardware and software, and telecommunications. These industries
present several additional challenges for antitrust enforcers. One is
that antitrust enforcers must promote both competition and innovation
in these fields through a balanced treatment of intellectual property.
Another is to account for the tendency for network externalities,
common in many high-technology fields, to create a strong potential
for market dominance. A third challenge is to anticipate future
developments in these fast-paced industries and conduct antitrust
policy accordingly.

INNOVATION AND INTELLECTUAL PROPERTY

The key assets in high-technology industries are often not factories
or machines but intangibles such as scientific ideas or the algorithms
contained in computer programs. These assets, unlike physical assets,
can be used by any number of people at once. Without intellectual
property protection, firms and individuals would have insufficient
incentive to produce these assets, because they are costly to produce
but cheap to copy or imitate. In recognition of this problem, the U.S.
Constitution empowers the Congress to ``promote the Progress of Science
and useful Arts, by securing for limited Times to Authors and
Inventors the exclusive Right to their respective Writings and
Discoveries.'' Patent and copyright laws do just that.

An important initiative of this Administration has been its use of
antitrust enforcement to further encourage innovation and to clarify
the role of intellectual property in antitrust law. The Administration
recognizes that the licensing of intellectual property for use by
persons other than its creator can benefit society both directly, by
allowing the more widespread use of intellectual properties, and
indirectly, by increasing the return to such assets and thereby
encouraging innovation. Such licenses, however, sometimes contain
restrictions that limit competition and actually discourage
innovation. These restrictions may violate the antitrust laws.

The recent case involving the British firm Pilkington plc, the world's
largest float glass producer, provides one example of the Justice
Department's attempts to use antitrust enforcement to encourage
innovation. Beginning in 1962, after acquiring hundreds of patents
worldwide on glass production processes, Pilkington entered into
licensing agreements with all of its principal competitors. These
agreements generally included territorial restrictions, so that each
licensee could construct and operate float glass plants in only one
country or group of countries. These restrictions allegedly limited
the incentives of Pilkington's competitors to innovate in glass
processing, by geographically restricting their opportunities to
exploit such innovations. Their incentive to innovate was allegedly
further limited by requirements to report any improvements in float
glass technology and to cede the rights to such improvements back to
Pilkington. In 1994 the Justice Department entered into a consent
decree with Pilkington, which, among other prohibitions, enjoined
Pilkington from enforcing its licensing restrictions against U.S.
licensees. The Justice Department's case was strengthened by the fact
that Pilkington's principal patents had expired long before the
complaint was filed. The Department does not, however, in general
limit its attention to restrictions that outlive the life of patents.

The 1995 Antitrust Guidelines for the Licensing of Intellectual
Property explain the balanced approach taken by the antitrust
agencies. The guidelines recognize that intellectual property
licensing can create efficiencies by allowing firms to combine
complementary factors of production. However, licensing arrangements
such as those used by Pilkington may contain restrictive terms that
reduce competition among alternative technologies, and the antitrust
agencies have sought to eliminate such anticompetitive arrangements.
In evaluating the licensing of intellectual property, the agencies
balance the procompetitive and anticompetitive effects.

NETWORK EXTERNALITIES

Many high-technology industries such as computers and communications
exhibit network externalities: that is, consumers derive more value
from the products of these industries the more people use them. For
example, a computer program often becomes more valuable as its network
of users grows, because users like to trade data files and exchange
ideas about how to use the program effectively. Network externalities
can sometimes therefore make entry difficult, because small firms may
be unable to compete effectively against large ones, whose products
enjoy additional value from widespread usage.

The challenge for antitrust policy is to preserve the benefits of
network externalities for consumers while preventing firms from
exploiting the market power to which these externalities can give
rise. When sellers agree to standards, consumers benefit because the
products of different sellers are then compatible. Unfortunately,
however, firms can sometimes manipulate the standards-setting process
to their own advantage, as the FTC claimed happened in a 1995 action
against Dell Computer Corp.

Dell was a member of the Video Electronics Standards Association
(VESA), a standards-setting organization in the computer industry. In
1992 VESA set a new standard for the design of computer bus hardware
(the hardware that transmits information between a computer's
components). According to the FTC, before the standard was approved,
Dell certified that it did not violate any of its intellectual
property rights, but after the standard was implemented the company
announced that the standard did violate one of its patents. Since by
then over a million computers using the standard had already been
sold, other computer manufacturers could not switch to an alternative
design without creating a compatibility problem. This would have put
Dell in a good position to collect substantial royalties on its
patent, were it not for a settlement with the FTC, in which Dell
agreed not to enforce its patent rights against computer manufacturers
using the standard.

FAST-PACED TECHNOLOGICAL CHANGE

The fast pace of change in high-technology industries makes it hard
for antitrust enforcers to anticipate the impact of future
developments when deciding the proper course of action. For example, a
merger that seems innocuous today may eliminate future competition.
Alternatively, a merger may increase concentration significantly today
but may not pose anticompetitive problems, either because of entry, as
discussed earlier, or because of exit, as revealed by the 1997 merger
between Boeing Co. and McDonnell Douglas Corp.

Although the Boeing-McDonnell Douglas merger reduced the number of
sellers of large commercial aircraft worldwide from three to two,
thereby sharply increasing concentration, the FTC decided that
McDonnell Douglas's 5-percent market share overstated the company's
likely future competitive significance, because this market share
reflected only the filling of old orders. Extensive interviews by the
FTC revealed that advances in aviation design had left McDonnell
Douglas behind: since the firm had not invested as much as its
competitors in improving the technology of its aircraft, the vast
majority of airlines no longer considered purchasing its aircraft. As
a result, the merger did not eliminate viable future competition in
the commercial aircraft market. Moreover, after consulting with the
Department of Defense, the FTC concluded that there were no prospects
for Boeing and McDonnell Douglas to bid on the same defense projects.
Having concluded that the merger raised antitrust concerns in neither
commercial nor defense markets, the FTC did not challenge the merger.

Future competition was a critical issue in the investigation of Bell
Atlantic Corp.'s 1997 acquisition of NYNEX Corp. The merger did not
increase current concentration in any local telephone market, because
neither Bell Atlantic nor NYNEX competed in each other's markets at
the time of the merger. However, the Justice Department and the
Federal Communications Commission (FCC) needed to assess the
likelihood that, in the absence of the merger, each company would
someday enter the other's geographic market, and the likely extent of
other firms' entry. One focus of the Justice Department's
investigation was the effect of the merger on future competition in
local service in New York City and nearby portions of NYNEX's service
area. NYNEX was the dominant supplier in that area, whereas Bell
Atlantic was one of many potential entrants.

After carefully studying the plans of other potential entrants, such
as AT&T Corp. and MCI Communications Corp., the Justice Department
concluded that the prospect for entry by a number of experienced,
capable, and well-financed competitors was significant. Therefore it
was by no means clear how much the loss of Bell Atlantic as an
independent competitive force would adversely affect consumers,
particularly given the evidence concerning efficiencies. The Justice
Department concluded that it could not meet its burden of proving that
the loss of Bell Atlantic as an independent entrant was likely to have
so significant a market impact as ``substantially to lessen
competition,'' the test of a violation under Section 7 of  the Clayton
Act.

The FCC, on the other hand, which also had authority to review the
Bell Atlantic-NYNEX merger, operates under a different statute with a
different substantive standard. Under the FCC's interpretations of the
Communications Act of 1934, the merging parties had the burden of
proving that the merger would on balance enhance competition and be in
the public interest. The FCC concluded that the merger would not
enhance competition, and it exercised its power to place conditions on
its approval of the merger. To remedy the merger's possibly
anticompetitive effects, and to advance the goal, set forth in the
Telecommunications Act of 1996, of opening local telephone markets to
competition, Bell Atlantic offered to make several market-opening
commitments, which the FCC accepted before approving the merger.

THE GLOBAL MARKETPLACE AND INTERNATIONAL ANTITRUST EFFORTS

The emergence of a global marketplace for many goods and services has
important implications for U.S. antitrust policy. On the one hand, as
transportation costs and trade barriers fall, many problems in
antitrust become easier. Mergers that would have led to significant
concentration in the absence of international trade may not do so once
one accounts for foreign competitors. Also, domestic price-fixing
agreements will be undermined if foreign competitors are willing to
sell in the U.S. market at a lower price. On the other hand,
international price-fixing agreements are more difficult than domestic
ones for U.S. antitrust enforcement agencies to police; success often
requires cooperation with foreign governments or international
organizations.

Unlike in the 1980s, when most antitrust fines were imposed in
domestic bid-rigging cases, the vast bulk of the over $200 million
imposed by the Justice Department's Antitrust Division during fiscal
1997 was collected in judgments against large international
price-fixing conspiracies. This suggests that even though
international trade may make price fixing more difficult, it will
probably remain a serious concern for some time to come.

Criminal prosecution in international price-fixing conspiracies is
generally much more difficult and complex than prosecuting domestic
conspiracies. First, the antitrust authorities must demonstrate that
U.S. antitrust law applies. In 1997 the Justice Department made
significant headway on this point, when the First Circuit Court of
Appeals held that Section 1 of the Sherman Act applies to ``wholly
foreign conduct which has an intended and substantial effect in the
United States,'' regardless of whether the case is civil or criminal.
Even when U.S. antitrust laws do apply, crucial evidence or culpable
individuals or firms may be located outside the United States and be
beyond the jurisdiction of U.S. courts.

These jurisdictional problems make it imperative that U.S. antitrust
enforcement authorities coordinate their activities and cooperate with
authorities abroad. In several recent investigations, the United
States made good use of its mutual legal assistance treaties with a
number of foreign countries: the Justice Department sought and
received assistance in cartel investigations from several countries,
including Japan and Canada.

In 1994 the Congress passed the International Antitrust Enforcement
Assistance Act (IAEAA), which empowered the U.S. antitrust enforcement
agencies to negotiate reciprocal agreements with foreign antitrust
enforcers. Under these agreements each government will assist the
other in obtaining evidence located in the country of the former,
while ensuring confidentiality. Unfortunately, foreign antitrust
authorities have been slow in following the U.S. lead in negotiating
these agreements, in many cases because they lack similar legislative
authorization from their own governments. In April 1997 the United
States nonetheless managed to negotiate its first proposed agreement
under the IAEAA, with Australia. The United States has also been
pursuing discussions with the Organization for Economic Cooperation
and Development toward a formal recommendation by that body that would
encourage its member countries to enter into mutual assistance
agreements that would permit more sharing of evidence with foreign
antitrust authorities.

At the same time the United States has also worked to improve
international antitrust enforcement through the so-called positive
comity approach. This approach is used in cases where markets outside
U.S. jurisdiction are affected by anticompetitive behavior that harms
U.S. interests. Under a positive comity agreement, if one country
believes that its firms are being excluded from another's markets by
the anticompetitive behavior of firms there, it will conduct a
preliminary analysis and then refer the matter to the foreign
antitrust authority for further investigation and, if appropriate,
prosecution. In April 1997 the Justice Department announced its first
formal request to the European Union under a 1991 positive comity
agreement. The Justice Department asked the Directorate General IV (DG
IV), the European Union's antitrust arm, to investigate possible
anticompetitive conduct by European airlines that may be preventing
U.S.-based computer reservation systems from competing effectively in
Europe. DG IV has announced that it is actively pursuing the matter.

Another notable ongoing effort in this domain is the competition
advocacy program undertaken jointly by the Justice Department and the
FTC. The two agencies are working together, in programs funded by the
U.S. Agency for International Development, to educate and otherwise
assist governments of developing countries in setting up antitrust
enforcement programs. This assistance has included helping countries
to draft competition laws, setting up implementation procedures,
training their staffs, and, in some countries, placing long-term U.S.
advisers in the antitrust office. Several countries in Eastern Europe
have benefited from this extensive interaction with the U.S. agencies,
and the program has now expanded into countries of the former Soviet
Union and Latin America.

Although significant progress has been made in international antitrust
enforcement, the growing importance of international trade makes it
imperative that the antitrust enforcement agencies continue their
efforts in this area. To this end, the Justice Department has
established the first-ever International Competition Policy Advisory
Committee, comprised of distinguished business, labor, academic,
economic, and legal experts, to advise it on these cutting-edge
issues. Investing in expanded enforcement and globalization of
antitrust principles will lead to better protection of competition
worldwide, and will yield substantial benefits that can be shared by
many.