Merger Policy and Schumpeter’s Creative, and Destructive, Gale

The organizers have advised me to craft a paper
that is designed to stimulate discussion; one even
advised me to es
chew footnotes and other scholarly
, which I have tried mightily but not
entirely successfully to
do. What I propose to do here
is avail myself of what I understand to be the virtues
of the discussion format

lay out thoughts and
problems an
d then wait for the discussants to clear
things up for me.
I might borrow from Alison Oldale’s
formulation and say that the views I express might not
even be my own

at least not after hearing your
comments. Since the audience includes many
who in the end decide these things, I will
pay careful attention to all
2The first matter worthy of consideration when we
reappraise current merger policy is the fundamental
question of market definition, a key to deciding
whether a merger
is or is not likely to have an adverse effect on competition.
Or some might say, in times
it was
such a key.
I have long had some doubts
about the wisdom of that exercise,
ever since
Professor Yarrow
raised the issue with me. Yarrow
now has an
increasing band of supporters. Louis
Kaplow, of Harvard University and the National
Bureau of Economic Research, recently made what
he characterizes as an “immodest claim” that:

1Alison Oldale, “Market definition is dead. Long live market definition?”, p.1. All papers referred to herein by author and title only were presented at this conference.
2 Indeed, this version of the paper replaces the original draft to incorporate some of the remarks at the conference.

the market definition process is incoherent as a matter of basic economic principles and should be abandoned entirely. This conclusion is based on the inability to make meaningful inferences of market power in redefined markets; … the impossibility of determining what market definition is best in a sensible manner without first formulating a best estimate of market power, rendering further analysis pointless and possibly leading to erroneous outcomes….4

An intriguing thought, but in the real world

according to Philip Lowe’s introductory remarks I am

3 In the following paragraphs I draw on a talk I recently gave at a conference sponsored by the Regulatory Policy Institute at Merton College, Oxford.
4 Louis Kaplow, “Why (Ever) Define Markets?”, The Harvard John M. Olin Discussion Paper Series, Discussion Paper No. 666, March 2010.

here to represent that world, apparently for lack of continuing qualifications to be classified as an academic — not entirely useful, for several reasons. First, as several participants point out, different national jurisdictions emphasize different tests, and the use of market definition remains attractive to several of them. Second , as Barry Hawk points out, there is no unanimity among economists and lawyers on this point and,
third , there is the small matter of the courts , which in America at least cannot ignore the “line of commerce” provision of the governing statute, which requires that the requisite lessening of competition be in a definable “line of commerce.” 5 That sounds to this economist as demanding some sort of market definition, perhaps for the use suggested by Ms. Oldale — as “part of the organising framework for the competitive assessment. ”6 In any event, it is clear from the 2010 Horizontal Merger

5 Barry Hawk, “A Tale Of Two Cities: Washington And Brussels Face The Courts”, pp. 7 and 9.
6 Oldale, p. 8. She adds “Market definition plays a role here in identifying the products of interest, identifying horizontal overlaps and potential non
– horizontal concerns, identifying those distinct markets requiring a separate analysis, focusing the process of evidence gathering and so on.”

Guidelines that
market definition
is still regarded by
regulators as an
analytical too
l.My guess is that we will end up with some
combination of reference to market definition

perhaps renamed to obscure the fact that the tool so
derided by so many is still in use

and the alphabet
soup of newer measurement devices discussed by
Fletcher, Alison Oldale and Chris Walters in
their paper, in which they are careful to point out that
“market shares and concentration measures… still
have a valuable role to play when accurately
brings me a worry and a suggestion

ave to pay less attention to static definitions of
market share and more to the durability of existing
in the face of a gale of creative destruction
. In
short, the question should not be
, “D
oes a company
have a dominant market share

or at least
should only be the start
ing point in an inquiry

but “I
that market share likely to prove ephemeral
. I

7 Amelia Fletcher, Alison Oldale and Chris Walters, “UPP: Up and away? The rising prevalence of pricing pressure indicators.”

suppose this is merely another way of asking some of the questions we have always asked: A re switching costs high? C an the incumbent make entry difficult and costly ? Does the incumbent have a history of anti competitive behavior? Is the nature of likely entry such that new entrants will merely nibble around the edges of the currently dominant firm, or is the incumbent more likely to be swept away in the famous perennial gale of creative destruction? Perhaps these questions are nothing more than Carl Shapiro has captured in his briefer formulation, that we have to recognize that “each industry has unique features” — an unfortunate fact for those lawyers who say they pine after certainty , which in the case of European merger control is declining , say Professor Duso and his colleagues. 8

If a seemingly dominant company in an
of concern to the enforcement authorities
does make
an acquisition,
subject to

assaults from new entrants sporting new technologies

8 Tomaso Duso, Klaus Gugler and Florian Szücs, Merger Policy Evaluation: Where Do We Stand?”, p.7.

and better mousetraps , we might well need a more careful analysis of market – share or market – power durability than
that in which we have traditionally engaged if we are to avoid Type 1 errors , which Professor Duso and his colleagues report are on the increase .9 As Commissioner Joaquín Almunia recently remarked, we must “allow competition policy to look forward on the basis of innovation and not only on prices and output.”10

If a seemingly dominant company in an industry of concern to the enforcement authorities does make an acquisition, and is subject to life – threatening assaults from new entrants sporting new technologies

Please understand: I am not saying that these
industries warrant special exemption, or inattention to
any anticompetitive practices in which they might
In a world in which intellectual capital is
rapidly replacing physical capital as the
of the wealth of nations, nothing can be more
important than
merger policy and, indeed, all
aspects of
the antitrust laws
with strict attention to the

competitive effect of certain trade practices, and the potential of any merge
r to affect competition

9 Op. cit., p.3.
10 Friends of Europe, “New Transatlantic Trends In Competition Policy”, Summer 2010, p. 18.
11 In the next section I draw on a talk I delivered to The Federalist Society for Law & Public Policy in Washington, D.C., on November 11, 1998.

But I
saying that
we might want to adopt a
merger policy that distinguishes between two cases,
in both of which the acquiring firm has a dominant
position, or is about to acquire one, as we traditionally
measure such
a position. In the first case, we are
dealing with a dominant firm that might have its
market share eroded a bit by new entrants: we can
safely use its current position as some indication of
dominance. In the second case, we might be dealing
with an acquis
on by a firm with a
dominant position
as traditionally measured,
but a position
subject to
destruction by a creative gale generated by a new
Are we to treat these two cases as if there
were no such difference?So I would
suggest for discussion the question of
whether what is required to accommodate the new,

12 See, for example, Amir Efrati, “Start Up Aims at Google”, The Wall Street Journal , November 1, 2010, the tale of Blekko, Inc.’s challenge to Google. Or consider the fate of Myspace, which ceded so much market share to upstart Facebook that at this writing its fate is uncertain, or of Kindle, a venture of Amazon that almost immediately found its dominant market share eroded by iPad, which in turn faces new competition from a reconstructed and repriced Kindle and other devices.

more rapid rate of
innovation that
characterizes many markets is a
shift in emphasis
from current to prospective market shares,
consideration not only of the current state of
play, but
what is looming on the horizon
. This is
in industries that are now coming in for
increased regulatory attention
d are included in
the list that
Almunia recently cited as
“high on our enforcement agenda”.
r two reasons.First,
these industries
a state of flux:
previously unimagined and
unknown businesses can
suddenly attract

billion users;
advertisers are
continuously calculating and recalculating the
marginal cost:marginal revenue relations
hips fro
various expenditures;
in some aspects competitors
are only a click away
and eager to serve consumers
In these instance
s it seems to me
that it behooves
regulators to think carefully before
concluding from

13 “New leaders on both sides of the Atlantic are showing renewed interest in scrutinizing innovators. Competition agencies ideally help consumers by ensuring open, competitive markets while eschewing actions that impede innovation and competition. But the accelerating pace of technological change makes their task more difficult.” Timothy J, Muris, “Antitrust in a High  Tech World,” The Wall Street Journal , August 12, 2010. One such firm, Google, is a client of this writer, and others that might have some interest in this topic, once but no longer clients, include BSkyB, News Corp, News International, and AMD.
14 “State of Play and Future Outlook, Competition Day”, Brussels, 21 October 2010.

measurements of
market share
, uncertain
guides in any event,
that a firm is dominant in any
sense that should be considered when appraising the
competitive effects of a proposed merger.Second, we find ourselves at a peculiar time. The
firms that we might for these purposes call high


the Googles, Microsofts, Apples, Twitters,

are in most cases awash in cash. If they
are to grow,
they will have to rely at least in part,
and perhaps in major part
on acquisitions, generally
of firms that, added to their existing products, will
enhance the package o
f services they offer to their
customers. Often, these service and product
enhancements make life easier for customers
Intel argues will be the result of its acquisition of
necessarily adversely
affecting the
ability of newcomers to enter any market. In such

15 Some investors would prefer share buy backs or higher dividends to growth by acquisition, but that is rarely a popular strategy with managers who have so far succeeded in producing double digit growth for shareholders, and are confident they can continue to do so, this time around by adding strategic acquisitions to continued organic growth.
16 See Don Clark, “Intel Looks to Protect More Than Computers,” The Wall Street Journal , August 20, 2010. Intel claims that the $7.7 billion acquisition (a 60% premium) will result in new chips that better protect PCs and other devices without slowing them down, and among other things allow consumers to remotely turn off stolen or lost cellphones. See also, Eric J. Savitz, “Intel’s Brilliant or Bonehead Deal, Barron’s, August 23, 2010.

cases it is important to consider the competitive
history of the acquiring company
a fact that might
weigh heavily on
in the future.

Note: any such enhancement of the consumer experience is different from claimed efficiencies. I note that Professor Röller, in his paper, reports that “efficiencies have not played much of a role in merger assessments”.18 Unfortunately, Professor Röller finds that alleged static efficiencies “were not infrequently accepted by DG Comp” at times while “alleged dynamic efficiencies have never been accepted.”19 I say “unfortunately” for two reasons. First, most studies have found, and this is confirmed by real world experience, that claimed reductions in “variable and marginal cost” — the static efficiencies most likely to be passed on to consumers — have very rarely been realized in practice . Second, it is dynamic efficiency that we are after — the driver of economic growth and improved living standards. The fact that

17 It seems obvious to me that Microsoft did not obtain its current market position merely by the exercise of “superior skill, foresight and industry,” to borrow from Judge Learned Hand’s Alcoa decision.
18 Lars Hendrik Röller, “Efficiencies in EU merger control: do they matter?”, p.1.
19 Loc. cit. , pp.6-7.

they cannot meet the verifiability, merger

specific and
consumer benefit tests is understandable: they are
not immediately realizable, and it takes something of
an act of faith to believe they will appear as a result of
a merger. Which brings
me back to my suggestion
that when working on the question of market definition
in merger cases, we can a
s a substitute for the well


effort to identify dynamic efficiencies
incorporate the possibility of dynamic efficiency gains
by giving
weight to the destructive potential of new
entry when defining “dominance”.

Of course, some would argue that the acquisitions I have been discussing represent vertical integration, and therefore cannot add to existing market power . I would
disagree . Although “such non – horizontal mergers are, in principle, less likely to create competition concerns than horizontal mergers”,20 as Dr. Lübking points out, in some instances vertical mergers, in his words “may significantly impede effective competition”, perhaps by

Dr. Johannes Lübking, “Some reflections on the application of the substantive test and of the substantive
merger guidelines,” p.6.adding
to the market power of the acquirer by making
entry by other than offerers of a similarly full line more
difficult, or
, put differently,
reducing the ease with
which small,
firms can carve out a niche for
On the other
hand, regulators generally,
or at least the European Commission recognize that
“vertical and conglomerate mergers provide
substantial scope for efficiencies…”.
My own guess
is that the great majority of the vertical and
conglomerate mergers we are likely
to see in the
industries that are attracting increasing attention

will when they dip into their substantial cash piles

will pass muster as providing considerable customer
benefits with little danger of creating barriers to entry
or enduring and unacceptable market power. Still, it
will take analysis of the specifics of each merger to
separate the pro

competitive and competition

from those that are likely to have significant anti
– competitive effects and, perhaps, to identify dynamic
Ibid.I raise these questions for discussion because I
find myself uncertain about appropriate policies to
deal with these difficult issues.
And because if we get
merger policy wrong, we will either prevent mergers
that might benefit consumers, or perm
it those that
confer exclusionary power on the merged entity,
an expensive effort to uncover and
force abandonment of those practices

with even
what seem the most draconian fines hardly making a
dent in the ill

gotten gains accruing from
behavior, although if commissioner Almunia succeeds
in setting “the cost of non

respect of competition
rules” higher than the projected gains from anti

competitive behavior
that situation will change.
23Do forgive me for the ambiguities in this paper.
They are a testimonial to my faith in this seminar to provide answers to difficult questions, answers that
have eluded me.

22 For the role of static and dynamic efficiencies in merger regulation see Röller, op.cit.
23 See Joaquín Almunia, talk at Competition Conference sponsored by Businesseurope and US Chamber of Commerce, 25 October 2010.