Merger Policy with 2020 Foresight: Efficiencies and Entry, Remedies and Retrospectives: Comprehensive Reforms for Comprehensive Needs

John Kwoka:

The second panel is concerned with several factors that
play important roles in merger review. These include the use and perhaps
overuse of entry and the efficiencies as defenses, the use of remedies as
challenges to mergers, issues of potential competition and innovation, and the
benefits of merger retrospectives. And to speak about these issues, we have
three people who are familiar to all of us. So let me just briefly update you
on their recent work. Bill Baer of course was the Assistant Attorney General
for Antitrust from 2013 to 2016. He’s also been a director of the Bureau of
Competition at the FTC, and is now a visiting fellow in governance studies at
Brookings. His list of well deserved honors will take more than half of our
time to list out, so I will postpone that for another occasion.

Michal Gal is an international scholar and director of the
Law and Markets Program at the University of Haifa, in Israel. She’s also
currently president of ASCOLA, the Academic Society for Competition Law. I
might mention the ASCOLA Conference will be online starting on June 25, is that
right, Michal? And I would urge you to register for that. Rich Gilbert is
professor of economics at Berkeley, He’s done path breaking work on innovation,
intellectual property and antitrust policy, both during his time as chief
economist at the Antitrust Division in the 1990s, and now with his forthcoming
book called Innovation Matters, Competition Policy for High Tech.

So we have a large number of topics potentially to cover
but fortunately we have 45 minutes to handle all ten of these, so there should
be no problem with our distinguished panel. Let me begin with you Bill if I may.
You’ve had of course long experience of both the agencies and in private
practice. My question to you concerns efficiencies. The early version of the
merger guidelines said that only extraordinary efficiencies would actually be
considered, that the numerical thresholds and the guidelines were designed to
accommodate routine modest efficiencies. Of course that standard, that guidance
has changed over time so that it’s a rare merger that doesn’t claim some
efficiency benefits from the merger. The business and economics literature has
continued to be skeptical of these claims. And so I wanted to ask, how
satisfied should we be with the way the agencies evaluate claims of
efficiencies as a defense? Are there changes that could usefully be made in the
way that agencies approach these issues?

Bill Baer:

Thanks John, and congratulations on the new book. It is
yet another great contribution to you to many of us and IO issues. Look, do the
agencies and maybe more importantly do the courts have the right level of
skepticism about efficiency claims? I think perhaps not, it’s so easy for
parties to predict that there will be enormous cost savings associated with a
particular transaction. And it’s very hard to establish the likelihood they’ll
be achieved and the merger specific nature of them. I was around at the FTC
with Bob Pitofsky in 1996. And so when he fought hard that we needed to find a
more systematic way to take efficiencies into account.

And I actually went back to my old law professor, Bill
Baxter, who was the predecessor of mine at DOJ, and talked to him about that,
visited him out of Stanford. He was – despite his Chicago School rigor and
commitment to that approach at Antitrust – he always thought looking at
efficiencies was a rabbit hole the government should not go down. That was
going to distract from making the tough prediction about whether there was a
likelihood of competitive harm. And I think looking back he may well have been
right, that the debate mostly in court is very often 50% about the nature and
extent of the efficiency claims. And I think that detracts from the
government’s effort in its burden to tell a credible story of competitive.

So I think we’ve got a couple of issues here. One is the
difficulty in determining whether or not the efficiency claims are likely to be
realized, and as merger guidelines recognize… The evidence about efficiencies
is largely in the hands of the merging parties, and it’s a little hard to drill
down and to penetrate that. To the extent, and you referenced this in your
opening, there is literature talking about realization of efficiency claims, I
think very often those claims are vastly overstated. And in the Staples Office
Depot matter years ago, 1996-1997, they came up with ridiculously trumped up
claims that the judge ultimately rejected.

But there’s another issue here that even to the extent
that some degree of cost savings ought to be credited in connection with the
mergers, there is this issue of whether or not they’re going to be passed
through, whether the market post-merger is sufficiently vigorously competitive
that consumers will benefit from those cost savings. And so you have a
potential situations in markets which today are or in the future may well be
less competitive than is ideal, that the cost savings are going to be won by
the shareholders of the merged entity. The consumers in effect will be paying
more than they would in but four world.

And indeed, another burden is going to be by the employees
whose jobs are lost. That very often is a huge part of the efficiencies claims
that are associated with the merger. And so between those two things, consumers
losing the benefit of the pass through in markets that are less than
competitive, and employees laid off, you’re having a huge wealth transfer from
consumers and from employees to shareholders. And that is a real worry to me.

Kwoka:

So would you have suggestions for how the guidelines or
the practice or the representation to courts can be changed in order to address
this imbalance and this inclination to give substantial weight to
inefficiencies? Is there a way of ratcheting this back?

Baer:

Looking back again this morning at the horizontal merger
guidelines and they don’t really get into this issue of likelihood of pass
through, and that may well be something we ought to focus on. And you could
even set up a hurdle, which is that not only do these claims need to be merger
specific, we need to have a high level of confidence that those savings will be
substantially competed down to consumers because that’s the benefit of the
transaction. So if you’ve got a merger that there’s some competitive risks, to
even get into the efficiencies debate without having a high level of confidence
that there will be pass through to consumers, and therefore benefits to making
markets more competitive. I think you might want to exclude them.

Kwoka:

So another suggestion, and then I’ll ask for Michal and
Rich to join in. Another idea of course is based on the following observation,
that is merging parties right now claim as much in the way of efficiencies as
they can get economists and other experts to attest to when they come into the
agencies, whether or not they’re plausible and really without any way of
determining whether they in fact come to pass or not. So the question is
whether there should be some mechanism for accountability for ex-post
determination of whether the efficiencies in fact arise. Should the companies
for example be required as a condition of merger approval to submit sufficient
data and information for two or three or four years that would allow the
agencies to check whether efficiencies have in fact materialized?

I would leave that question out there because I know that
someone will say, “What if they find out that they haven’t materialized?
What do you do next?” But I think it would be at least informative and
maybe even a discipline on companies if they were obliged to report
sufficiently in order that the agency and arguably the broader public could
determine whether the efficiencies have in fact appeared. So I don’t know, if
you Bill want to comment on that and then I’ll ask Michal and Rich to join in.

Baer:

Sure, real quickly, and then I’ll surrender the microphone
and the video, which probably is beneficial to everyone. I do think having some
ability to get post-merger information about efficiency realization is a
worthwhile thing. And there may be a little bit of specific deterrence in that.
There may be little one can do to remedy it if the efficiencies aren’t
realized. But if you build a body of data that helps us apply an appropriate
level of skepticism to these claims, you’re going to affect the agency’s
ability to convince the courts to avoid letting the sideshow be the main event.

Kwoka:

Michal, do you have any observations I’m sure you do,
please offer some observations on any of this so far.

Michal Gal:

Okay, great. I mean, I’ll be very short. But a few
questions or issues that arise here in my mind, because for me it was very
surprising that while in theory efficiencies play an important role in mergers,
in practice, you can’t find many cases in which efficiencies have indeed
changed the outcome. And I actually looked. When I wrote my book a few years
ago, I actually looked and the closest I got was the, but that was overturned
by court. And so I asked myself, what’s happening here? Why do we see this in theory
but not in practice?

And so a few weeks ago I heard an interesting lecture by
Lewis Kalow from Harvard, who says that actually we cannot and we should not
separate the efficiency analysis from that of anti-competitive effects. And
what he says is that such effects actually come together, and that got me
thinking, and I think that I agree with it at least partly because we take
efficiencies into account if they strengthen the smaller player, thereby
strengthening the equality. Okay, so that’s where we’re into [inaudible
00:12:47] versus the true value [inaudible 00:12:54] has been significantly
limited competition.

Again, we want to know if it’s going to be able to limit
competition. So in order to answer these questions, we only need to take account
of efficiencies to see. I think we already have that to some degree in the law.
So that’s the first point. Another point that was made which I thought was
interesting, and I thought that we should bring it here is that we should look
at the theory of the firm and consider how firms operate internally. Okay, in
order to verify whether they can actually be efficient. So let me give you a
simplified example. Let’s assume that the firm has merged with many others, and
it always fails. I mean, can we take seriously the efficiencies that it saves?
So I thought these were interesting points.

Kwoka:

And Rich, do you want to comment on any of this?

Richard Gilbert:

Sure, John. Thanks. So Bill mentioned the importance of distinguishing
between the evidence about efficiencies versus the representations about
efficiencies. The evidence is a little weak, shall we say. I recall from my
experience at DOJ a long series of mergers in the hospital industry, which was
unfortunately one of the real – shall we say – failings of antitrust
enforcement. It led to a lot of concentration, a lot of problems in my opinion
for patients. Now, in each one of these mergers the parties came forward with a
really long and detailed list of efficiencies that were going to happen from
these mergers, and you look at it, they look pretty credible. The problem is
that they weren’t implemented.

The efficiencies were things like closing one of two
surgery centers because they were redundant. Of course, the doctors like their
surgery centers and didn’t want to close after the fact. Now, what we could
have done, we meaning antitrust enforces could have done as condition approval
of these mergers or give a grace period during which these claim efficiencies had
to be actually put into effect, otherwise the merger would not be allowed to be
consummated. But that didn’t happen. I think it led to a lot of problems in
that industry.

Kwoka:

So let me toss out yet another difficulty that I’ve
observed I think in my examination of some mergers, and that is that so called
efficiency claims seem to have changed. The traditional efficiency claims had
to do with changes in marginal costs. The economists and attorneys and
accountants and others had eventually figured out ways of doing a pretty good
job at estimating those types of cost changes and traditional efficiencies for
mergers. But more recently, parties claim benefits not really efficiencies, but
take the form of greater quality, or enhanced services, or more investment or
faster rollout of something, or as I’ll get to later with Rich, innovation.
These benefits are to say the obvious a lot harder to measure, and they’re
harder to either prove if the burden falls on the parties or to disprove if the
agencies need to rebut them.

So my question now is, as the focus of claimed benefits
from mergers – as it has shifted to these – added yet another burden on the
agencies to deal, grapple with things for which not even economics has answers
that are straightforward enough to implement, that we moved really into the
gray area, into the deeper darkness area of trying to assess the benefits from
mergers. Does anyone have experience or comments or reactions to that concern?

Gilbert:

Well, I’m happy to comment on that because it’s matching
the innovation. I’m not convinced that these types of claims are more difficult
to evaluate than other claims. I mean I gave the hospital example where you had
all of these very specific operational claims it’s just that they weren’t put
into effect. Now we do see some mergers where I think the parties perhaps do
have credible claims about how there are complimentary effects. And those
benefits should in fact be given weight if they are credible and merger
specific. So my view is not that we should discount these claims, but rather we
should put them to the test where they’re actually credible claims that are
going to be implemented. They don’t necessarily have to be marginal cost based
claims, they can be claims that are, as you say, pertinent to quality or the
rollout of new products and services. And if they’re real, go for it, if
they’re made up, don’t.

Kwoka:

Michal, what’s your view from international experience,
either in Israel or other countries where you’re familiar with? How did they
deal with efficiency arguments more generally?

Gal:

Well, I think that efficiencies are very hard to prove
regardless of the jurisdiction, and especially if the burden there is quite
high. However, what I see from my work on small economies is that there is
often more flexibility if I can see it this way, and there’s a good reason for
it, because think about a large economy which in most of its markets has quite
a large number of firms. So it can actually be more strict with the
requirements of efficiency and with its structural presumptions, whereas in
small economy, let’s say you have two firms or three firms and they are
inefficient, they are simply inefficient, and the costs are extremely high, and
many times they would need to join together in some way in order to become
efficient, in order to enjoy economies of scale, economies of scope, because
the size of the population is so low. So I think that in small economies making
sure that the burden is not high, is even more important than in large
economies.

Kwoka:

Good thanks. Oh, unless someone else has a something
urgent that they feel the need to impart to us I’m going to move on and ask
actually Michal the next question as well. I’d like to turn to the issue of the
importance of preserving entry and the concerns over mergers that eliminate
potential entrants. These seem, certainly in the US to be treated more
leniently than mergers between incumbents, and certainly there are reasons why
that may be the case, most obviously if the inside incumbent firm and the
outside by definition are not producing the same product, our tools of merger
analysis are an awkward fit to understand the potential competitive
consequences, and the agency’s must rely more on evidence about business
strategies and documents involving intent, and the courts are often not
particularly impressed by or convinced by that type of evidence.

And so potential competition mergers seem to be clearing a
lower bar. And my question to you and to all the panel is whether this comports
with your understanding and whether more particularly looking forward whether
there are methods and criteria for merger enforcement that would do a better
job of identifying potential mergers that eliminate potential competitors in
advance so that we could restructure and reform our enforcement practices with
respect to potential competitors to strengthen that side of policy. Michal, do
you have some thoughts you can offer us on that, please?

Gal:

Sure. And you’ve asked before if we have something urgent
to say, so actually do. I want to congratulate you on your book. Once again a
wonderful and thoughtful book that adds to the thinking in this area, so again,
I mean, thank you for it. And now for the issue of the day, you asked a very important
question, which is given our imperfect information and entry barriers and the
error costs involved in the analysis, on what side do we err? Okay, so let me
use what everybody calls the killer acquisitions. So just a metaphor on this.
Do we err on the side of not killing the killer acquisitions, or do we err on
the side maybe, using the same metaphor, potentially killing some of the
innovation markets?

Okay, so I think that’s part of the trade-off, and this
trade-off results because dominant exit strategy for many startups is often to
be acquired by the big ones. And this is partly because of the short term
interests of venture capital firms who invest in the startups want to see
immediate results, or quite fast results. So given the survival rate statistics
of startups, we want funding to be available, and so this is an important
consideration. Another consideration to take into account here I think comes
from another dimension, and come from studies of areas which have clusters of
high tech companies like [inaudible 00:24:29] and these study shows that
innovation oriented clusters create important positive spillovers on other
different players.

So these are important considerations to take into
account. However, it’s extremely important to make sure that we don’t kill
these types of technologies that can disrupt markets significantly. So let me
make a few points here. So the first point is legalistic, is that if we do want
to prohibit these mergers we may have to make sure that we can review them,
because some of the competition laws, or most of the competition laws around
the world have a single minded focus on sales revenues as the share for merger
review. However, this is problematic with regard to a technology that has not
been put to the market test yet. It might be disruptive, it might be great, but
it hasn’t been put to the market test yet. If you only look at sales, you get
zero.

And so, Germany and Austria have solved this problem by
looking at the height of the payment. How much did the acquirer pay for the
deal? And they set a threshold. So this is one way. Another way is certainly to
adopt a rule like in Israel, in Israel if you’re beyond 50% market share, you
have to report all your deal. Another way which was suggested by Mark Lindley
and Andrew McQueary, you focus on the acquirer, and require some kinds of
acquirers to come forward with these deals. So this is the first point, make
sure you can actually review these mergers. Second point is that much depends
on the technology, because in this regard I think it’s important to ask where
the competition comes from, especially with database markets, what we see is
that engineers are very high in many of these markets, so the chance of
introducing competition is by new technology that could circumvent existing
ones.

If you want, Schumpeterian competition might be a better
tool than displacement by competition over the same product. So some cases are
going to be easy case for the authorities in this regard. I can think about
some examples of new technologies in which displaces the incumbents in
innovative industries. Just to give you an example, let’s assume that firms can
create an algorithm that will requires much data points in order to make

[inaudible 00:27:51]

or it uses the same amount of data and makes much stronger
and trustworthy predictions. That algorithm can overcome some of the problems
of access to data that some firms have. This is disruptive. Okay? We do not
want certain mergers to go through.

However, sometimes a technology might be disruptive and
you won’t be able to, or the markets might not see its disruption before it
happens. And one of the examples that everybody uses is, of course Facebook and
Instagram. But I think another example, which I think is interesting is the
example of Twitter. And let me give you a disclaimer here that I’ve never used
it. However, from what I understand, it entered, it didn’t have a lot of data
because it didn’t have a lot of users in the beginning, but it’s offered
something that users like, it’s the sliding of the choices. Okay? So are
agencies really to know if this is going to be a technology which is going to
be disruptful, or that is going… I think many times the answer is no, you
wouldn’t know. So that makes this extremely hard.

Another question in this regard is, what… More difficult
cases pertain incremental change, or that firm set of features that cannot be
standalone or has so much benefits when it’s standalone. And then the question
is, what can we do about it? We can talk about it later, but I think these are
issues that we have to think about more. A third point is that we have to be
really realistic, and we have to engage in a holistic analysis of entry
barriers in such markets. And the idea here is that some of the decisions that
I read from all around the world, especially with regard to data markets are
not based on sound economics with regard to… Even what we know, there are many
things we don’t know, but some of the decisions are so simplistic that you say,
“Well, it’s quite problematic.” Do I have time to give an example?
Only if we have time.

Kwoka:

Please do, go ahead. You’ve set out a whole series of good
points by all means-

Gal:

Okay.

Kwoka:

… give us an example.

Gal:

Thank you. So some of the example I’ve been working with
in my own academic work, and it’s a joint venture, but the same things you see
in merger analysis, and this is a joint venture between Google and
Sanofi-Aventis, Sanofi is a global bio-pharmaceutical firm. And what they did
is they created a joint venture for the three important one, for an online
diabetes clinic, where data from its patients that it collects and Google
assists with the analysis and with providing predictions and suggestions to
patients in real time.

And so there were several issues with this joint venture
that the Commission, the European Commission looked at. And one for example was
the issue of interoperability. Okay. And the Commission said, “Well, the
joint venture has a very strong incentive for all kinds of products
interoperated to it because it will then have more data.” Well, it’s
definitely true, but on the other hand, if the interoperability is two
directional, the flow of information from this interoperability is one
directional, it goes only to the joint venture. This is a point that was not
raised by the Commission.

And then the other one has to do with a collection of
data. And here we’re going to make a very short point. I think it’s very
important. I think that the GDPR changed many things. It changed even for
American firms, it changed them partly because the American firms are
international and they work in Europe as well, so they’ve changed some of the
things. It changes because the GDPR is affecting legislators around the world.
You can see for example the California Act. And is pleased with regard to
private data. We have written, and for good reason maybe, I’m not saying that
it’s not for good reason. But we have created a much higher barrier for access
to private data. We have strengthened dominant firms. So this has to be taken
into account by the authorities when they analyze mergers. So I’ll stop here.

Kwoka:

Well, you’ve raised enough issues for us to spend another
three hours on, if not three years. So thanks for that. Really, no, you raised
a series of things. Let me get Bill and Rich back in on this and without my
interjecting anything here to see if there’s some parts of that, Bill, that you
would like to react to and offer please?

Baer:

Sure. A couple of points, first under US law the ability
to address acquisition of potential competitors or complimentary firms by a
dominant firm is quite limited, right? Under the Clayton Act you have to show
the firm that’s being acquired is one of the few actual potential entrants.
That’s a high barrier, it’s not going to be met with dominant firms under
current law, so the extent you want to deal with that you’re going to have to
tweak the Clayton Act, or resort to Sherman Section II, Monopolization
Enforcement, which is so far down the road it isn’t going to have any practical
effect.

So that’s point one. You would need to do what Michal
talks about, I think, in saying at some certain level of market share, maybe
persistent market share where you have network effects and tipping points, the
standards in terms of government intervention to block is low. But the second
point is that this is always framed up in the binary that startups won’t start
unless they know they can be purchased by Google, Apple, Amazon, Facebook. And
that’s really a false choice, right?

They won’t make as much money because in many cases there
is a market power premium being paid by the firm that is already dominant in
this market, and it wants to make sure it gets ahold of the innovators who are
potentially in that space or adjacent space, you would still have investors I
think willing to buy up good ideas, and they potentially then could market
those ideas to people who could combine and take on the dominance of the
existing incumbent. So we got to be careful about appearing saying things are
binary, there will be investment or there won’t, there will be startup or they
won’t, when in fact it’s much more of a continuum. Rich?

Gilbert:

Yeah, lots of very good points. I’m sure we’ll talk more
about it. And thanks Michal, very good discussion. But I would disagree in one
area here, which is, I don’t think in many cases, or at least not in all cases,
that the relevant trade-off is to kill or not to kill. You go back to the
killer acquisitions paper, Colin Cunningham and her co-authors identified that
the real problems occur when the acquisition has an overlap in either a
technology space or a product space. And that is typically where the agencies
have been pretty good about stopping those transactions, at least the
transactions that they know about. I want to give one little example, I’ll try
to be brief. So I consulted with the FTC on the proposed merger of Thoratec and
Heartware. These were companies that made a product you don’t want to
experiment with, it’s left ventricular assist device. It’s a heart pump.

And Thoratec was the major dominant really supplier of
heart pumps and Heartware was really an up-start. Now, the FTC, I believe,
properly blocked that merger on innovation grounds, because of the technology
as well as product overlaps. And here’s an interesting little anecdote, which
is a year or two later, Heartware sold out to a different company that was not
a competitor for something like four times the price that they were getting
from Thoratec. So everybody came out ahead in that transaction. So it’s not
always a trade-off. Just want to keep that in mind.

Kwoka:

As you rightly say Rich… Michal, do you want to respond
briefly?

Gal:

Sure. Just a quick word, I think I definitely agree. It’s
not a binary choice, and I agree with Bill as well that you can be bought for
less. And maybe there might be a combination of several startups together. And
I think that’s part of it, a part of the solution, for young competition law. We’re
asking ourselves whether we can create a market for a different type of
incentives for longer term incentives than these venture capital firms creates.
So it goes beyond merger policy, it goes to public policy. And I know that in
Israel for example, this is something we’re trying to do because it’s a startup
nation, but that’s what we have. We have mostly startups. We don’t have the
large firms. And so this can be part of public policy that can go hand in hand
with merger policy.

Gilbert:

May I add just one thing, which is that the latest
buzzword is now the reverse killer acquisition, the idea that these
acquisitions are preventing innovation by the acquiring company, as opposed to
innovation by the acquired company. I think we’ll be talking about that as well
at some point, probably not today, but it raises more complicated issues.

Kwoka:

Well, the four of us need to have a further offline
conversation about several of these things, for sure. But since we’re down to
10 or so minutes, I do want to turn to you Rich, and I do want to return to the
issue of innovation. It seems to me that the economics that we’re seeing the
wisdom from theory and empirical work in economics with respect to the effect
of mergers on innovation is it has less predictability than with respect to
let’s say, price or other traditional metrics. I think you may disagree, but
I’d like to hear you on that. And I think in particular, the innovation issue
dovetails with potential competition, if, if innovation efforts on the part of
companies are close to fruition then it’s possible to look at the effect of the
merger through the lens of potential competition.

But as one backs it up, so as, so it’s not commercialization
or development, but really primary research, then it becomes ever harder to see
what overlap there may be and what the risks may be from a merger that takes
one of the players off the board two or five or seven years down the road. And
so it seems to me that for both of those reasons, the challenges of dealing
with innovation effects of mergers really are significantly more difficult than
the hard challenge already of looking at price or output or some quality
outcomes. Is that a fair assessment, or is that too hard on the advances that
you and others have made in the innovation area?

Gilbert:

I believe it’s fair. And it’s worse than you say. Tougher
than you say. So we do have a lot of good studies. I want to applaud your book
and your work on retrospectives. Those are price retrospectives mostly, we have
almost nothing on innovation, that’s a big problem. And we have some theory,
good theory, just not enough empirical work and almost no retrospectives, at
least none conducted by the agencies. Now, the other area and Bill touched on
this that’s a major concern is suppose that the economists and the lawyers and
the enforcers get all of this stuff right, and we know what we’re going to do.
And in fact, I do think that the agencies have actually done a good job on
innovation, particularly in the pharmaceutical and related areas, but other
areas as well.

But now suppose you’re going to take this to court. How
are you going to make the case in court? I think we have such serious problems
in the courts because they have focused on procedures and tests and evaluations
that are just pretty well useless for innovation. And it starts with potential
competition. So I want to go back to potential competition a little bit because
you could think of innovation cases as harder potential competition cases. So I
want to talk about a case that we have heard about today a fair amount, which
is Sabre-Farelogix case, which in many respects is a potential competition
case, you had Sabre which has these global distribution systems, two sided
platforms to link airlines and travelers and travel agents. And then you have
Farelogix, which is really a software supplier to provide IT services to
connect airlines and travel agents.

Now, so DOJ challenged the merger, it went to court, I did
not work on this merger, I’m not party to any of the details, but I think
reading from the US courts decision is entertaining in a morbid way, which is I
want to quote a few things, which is, “The court found the following
facts. Number one, Farelogix has historically been an innovator, while Saber
has resisted change. Number two, Farelogix was an industry disrupter. Number
three, Saber and Farelogix view each others as competitors, though only in a
limited fashion. And number four, Farelogix has put downward pressure on
Saber’s prices.”

Now, you would think, again, I’m not part of the details
and there were some other issues that are relevant, you would think, okay,
we’re done. What else do we have to do? But wait, we haven’t done market
definition. And the court criticized the DOJ’s market definition based on
leaving certain things out, and the fact that they didn’t do a two sided market
analysis, and that Sabre only operates on one side of a two sided market. I
found this a little astonishing. In fact, it’s a little bit like saying, based
on market definition, the bumblebee can’t fly. I mean, you have the facts, what
more do you need? Now, of course, three days later, the UK Competition and
Markets Authority blocked the merger, and it specifically focused on the
innovation and it did not emphasize market definition or at least nearly not to
the same extent as the US Court.

So if we’re going to do innovation cases, they aren’t hard
cases, but if we’re stuck in the market definition box for not just for the,
shall we say, hard-to-understand conclusions that come out of some of these
analysis where we have facts but somehow the market definition says they’re
wrong. But also for innovation you have products that don’t yet exist, how do
you define a market for products that don’t yet exists, or for products that
are very early in the stage of development? Now, I think that should not be the
end all of the analysis, but at the core, I’m afraid it could very well be.

Kwoka:

So we all have our favorite recent cases that we like to
quote to our students so that whether in law school or economics departments,
so we can test our students understanding of the flaws here, but you do raise
an important point. Certainly in the US cases go through the judiciary and the
lack of education and understanding of modern economics and modern industrial
organization in the courts really has proven over and over again to lead to bad
judgments, to bad precedents, and really to a distortion in the cases that the
agencies bring because they have to thread their way through the likely
minefield of whatever court will ultimately review these.

So all of that is another problem that we haven’t yet
addressed, but one that is certainly relevant as we think about other reforms. They
are non-starters if they cannot be explained adequately and successfully to the
courts. You also raise an issue which I would like to… We’re running out of
time, but also just to touch on, and that is the role of retrospectives which
I’ve done some work on this and others have too which have proven I think to be
valuable and largely unexploited way of advancing our understanding about what
constitutes anti-competitive outcomes from certain types of mergers or
remedies.

And as you say, there are precious few of them in the
innovation area, and that certainly is a place that I think deserves much
greater attention from economists and the agencies as well. But we are running
out of time, I’d like to see…. Well, I, and I’m sure all of us have many
other things on our minds, I’d like to circle back and see if you Bill or
Michal have reactions to what Rich has just said before I ask for any final
comments you may have as well. So, Bill.

Baer:

I think Rich and I think we all are pretty much aligned on
the issue of nascent competition. It’ll be very interesting to see what the FCC
and Marble’s 60 Study turns up in terms of acquisitions by tech platforms, and
what the motivation behind was and what the outcomes were. I look forward to
that very, very much. And that is a retrospective study and I think doing more
of that as we talked about at the front end, is actually a very good thing.

Kwoka:

Maybe if I can just interject on that idea, said I look
back on what the agencies knew at the time that they made certain decisions,
what did they know at the time from the public evidence at least, about
Facebook and Instagram, Facebook and WhatsApp, and one of the most revealing
was Google Double Click, which the FTC reviewed and there’s quite a fascinating
dissent by one of the FTC commissioners that more or less got it right,
anticipated all the ways that Google is now being alleged to have seized
control and dominated the ad and ad servicing market. It’s not just with
hindsight some people had figured it out at the time. And even that’s worth
understanding why and how they did and others did not agree, the majority of
the FTC at the time did not agree. But Michal, anything further? Please.

Gal:

Yeah. I completely agree that retrospectives are extremely
important. I want to add here the political economy dimension. And I think that
you have to be quite a strong agency in order to engage in self-criticism and
engage in real one if it’s going to be published. So part of it has to do with
a question of how you choose your cases. If you’re a weak agency, you’re going
to choose the ones that you know the outcome and you come out good. So,
something to think about.

Kwoka:

So my timer has gone off, and they’ll probably pull the
plug on us soon enough. But again, if anyone has any final comments or
questions or reactions or suggestions, we’ve tried to make this to some degree
forward looking with ideas for where it is antitrust and merger policy in
particular should be looking to modify its approaches and strengthen
enforcement.

Gilbert:

I’d like to try to interject if we have a moment.

Kwoka:

For you we can.

Gilbert:

I believe that the approach that we currently have to
potential competition acquisitions, which is highly related to innovation
issues is just almost backwards. We focus on the likelihood that the target is
going to be realized as an actual competitor and how long that going to take?
To me These are irrelevant issues, if there’s one chance in 20 that a target is
going to mature into a real competitor, and if there are no offsetting
efficiencies, why do we want to take that chance? So the focus should be on,
what are the efficiencies from the transaction, if they are really credible and
merger specific, then the probabilities matter. But if there are no
efficiencies, the comparison is a one state in which nothing happens and
another state in which there’s a burn up, well, that’s not a type one type two
comparison. There’s only a bad thing that can happen. And we need to somehow
steer the analysis in that direction.

Kwoka:

So in any case let me with that bring our discussion to a
close and let me thank Rich again and Michal and Bill for a truly interesting
discussion. But let me also in concluding this event thank the first panelists,
Shawn, Larry and Aviv, with Diana moderating as well, certainly as Fred and
Steve for their very lively chat. I think and hope this event has made clear
how many areas of merger policy would benefit from further thought, from
further research and from some reforms. And it’s also I think highlighted,
brought to the surface a number of constructive and creative ideas for what the
necessary reforms might well look like. That’s been our purpose today,
foresight and thinking about future reforms, and we hope that everyone who is
listening in has found this discussion interesting and provocative. So thank
all the panelists and thank the audience for your interests. On behalf of
everyone here, stay safe, and for now, goodbye.