"The Apple iPod iTunes Anti-Trust Litigation"

Filing 753

***ERRONEOUS ENTRY, PLEASE REFER TO DOCUMENT NO. 754 *** EXHIBITS re 752 Opposition/Response to Motion, filed byApple Inc.. (Attachments: # 1 Exhibit 2, # 2 Exhibit 3, # 3 Exhibit 4, # 4 Exhibit 5, # 5 Exhibit 6, # 6 Exhibit 7, # 7 Exhibit 8, # 8 Exhibit 9, # 9 Exhibit 11, # 10 Exhibit 12, # 11 Proposed Order)(Related document(s) 752 ) (Kiernan, David) (Filed on 1/14/2014) Modified on 1/14/2014 (jlmS, COURT STAFF).

Exhibit 1

Page 233

Page 234

·1· · · · · · · · ·UNITED STATES DISTRICT COURT

·1· · · · · · · · · ·A P P E A R A N C E S

·2· · · · · · · ·NORTHERN DISTRICT OF CALIFORNIA

·2
·3· For the Plaintiffs:· · Bonny Sweeney, Esq.

·3· · · · · · · · · · · OAKLAND DIVISION

· · · · · · · · · · · · · ·ROBBINS GELLER RUDMAN & DOWD, LLP

·4

·4· · · · · · · · · · · · ·655 West Broadway

· · THE APPLE iPOD iTUNES· · · · · )· · Lead Case No. C 05-00037

· · · · · · · · · · · · · ·Suite 1900

·5· ANTI-TRUST LITIGATION· · · · · )

·5· · · · · · · · · · · · ·San Diego, CA· 92101
· · · · · · · · · · · · · ·619.231.1058

·6· · · · · · · · · · · · · · · · ·)

·6· · · · · · · · · · · · ·[e-mail]

·7· ____________________________· ·)

·7

·8· This Document Relates To:· · · )

·8

·9· ALL ACTIONS· · · · · · · · · · )

·9

10· · · · · · · · · · · · · · · · ·)

· · For the Defendant Apple, Inc.:
10· · · · · · · · · · · · ·David C. Kiernan, Esq.

11· ____________________________· ·)

· · · · · · · · · · · · · ·JONES DAY

12

11· · · · · · · · · · · · ·555 California Street

13

· · · · · · · · · · · · · ·26th Floor

14

12· · · · · · · · · · · · ·San Francisco, CA· 94104
· · · · · · · · · · · · · ·415.626.3939

15· · · · · · ·CONFIDENTIAL - ATTORNEYS' EYES ONLY
16· · · · VIDEOTAPED DEPOSITION OF KEVIN M. MURPHY, PH.D.

13· · · · · · · · · · · · ·[e-mail]
14

17· · · · · · · · · · · · VOLUME II

15

18· · · · · · · · · · ·January 08, 2014

16· Also Present:· · · · · Thomas C. Tracy, videographer
17

19· · · · · · · · · · ·Phoenix, Arizona

18

20

19

21

20

22· Reported By:

21

23· Cathy A. Miccolis

22
23

24· RPR, CRR, CSR No. 50068

24

25· Job No. 10009198

25

Page 235
·1· · · · · · · · · · · · ·I N D E X
·2· Witness· · · · · · · · · · · · · · · · · · · · · · ·Page
·3· · · ·KEVIN M. MURPHY, Ph.D.
·4
· · · · · · ·EXAMINATION BY MS. SWEENEY· · · · · · · · · 237
·5
·6
·7
·8· · · · · · · · · · · E X H I B I T S
·9· Exhibit· · ·Description· · · · · · · · · · · · · · ·Page
10· Exhibit 6· ·Supplemental Report· · · · · · · · · · · 257
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

Page 236

·1·
·2·
·3·
·4·
·5·
·6·
·7
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · · · ·THE VIDEOTAPED DEPOSITION OF KEVIN M. MURPHY,
Ph.D., VOLUME II, was continued on January 8, 2014,
commencing at 9:11 a.m. at the offices of BONNETT,
FAIRBOURN, FRIEDMAN & BALINT, P.C., 2325 East Camelback
Road, Suite 300, Phoenix, Arizona, before CATHY MICCOLIS,
a Certified Reporter in the State of Arizona.
· · · · · ·THE VIDEOGRAPHER:· The time on the record is
9:11 a.m.· Today's date is January 8, 2014.· My name is
Tom Tracy of Aptus Court Reporting.· The court reporter is
Cathy Miccolis of Aptus Court Reporting located at 600
West Broadway, Suite 300, San Diego, California 92101.
· · · · · ·This begins the videotaped deposition of Kevin
Murphy, Volume II, testifying in the matter of the Apple
iPod iTunes Trust (sic) Litigation, pending in the
District Court of California, Oakland Division, Case
Number C 05-00037 YGR.· This deposition is taking place at
2325 East Camelback, Suite 300, Phoenix, Arizona 85016.
· · · · · ·Will counsel please identify themselves,
starting with the plaintiffs' counsel.
· · · · · ·MS. SWEENEY:· Bonny Sweeney for the plaintiffs.
· · · · · ·MR. KIERNAN:· David Kiernan for Defendant
Apple, and Scott Murray, in-house counsel from Apple, may
be on the phone.
· · · · · ·Scott, are you on the phone?

Page 237

·1·
·2·
·3·
·4
·5·
·6·
·7·
·8·
·9
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · · · ·Okay.· No answer.
· · · · · ·THE VIDEOGRAPHER:· Thank you, Counsel.· The
court reporter may swear in the witness so we can proceed.

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · Q.· ·Okay.· And for the -- and we will just refer to
that as the supplemental Murphy and Topel report, if
that's okay with you.
· · · A.· ·That's fine.
· · · Q.· ·So with respect to the supplemental Murphy and
Topel report, did both you and Professor Topel write the
report?
· · · A.· ·Yes, we worked on it together.
· · · Q.· ·And how does that work when two people are
writing a single document in your case, how does that
work?
· · · A.· ·We first talk about what it is we want to do
and what -- you know, what we are, what we are presenting
in the report.· We then set out to write that up.
Obviously one person at a time writes.· We don't sit next
to each other and write.· So one of us will take a
different part and work on writing that part up.· And then
the other one will review it, make changes, edits,
whatever, probably discuss it further back and forth
between us with writing and discussing material that's in
the report.
· · · Q.· ·And you said that "each of us would take
different parts."· What part or parts did you have primary
responsibility for?
· · · A.· ·You know, I don't really recall.· I don't

· · · · · · · · · KEVIN M. MURPHY, Ph.D.,
having been first duly sworn to tell the truth, the whole
truth, and nothing but the truth, was examined and
testified as follows:
· · · · · · · · · · · · EXAMINATION
BY MS. SWEENEY:
· · · Q.· ·Good morning, Professor Murphy.· We have met
before.· My name is Bonny Sweeney.· I'm going to be taking
your deposition again today in the Apple case.
· · · · · ·MS. SWEENEY:· Before we get started, I just
wanted to ask counsel for Apple a question.· So I just
want to make sure, you have an open telephone line. I
want to know anyone who is on that line.
· · · · · ·MR. KIERNAN:· There is no one else on the line.
· · · · · ·MS. SWEENEY:· Okay.
· · · · · ·MR. KIERNAN:· It's just an open line for Scott
Murray, counsel from Apple, and if he joins, he will
announce himself.
· · · · · ·MS. SWEENEY:· Will there be any indication as
to when and who joins?

Page 238

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · · · ·MR. KIERNAN:· There is a -- it makes a beeping
sound.
· · · · · ·MS. SWEENEY:· Okay.
BY MS. SWEENEY:
· · · Q.· ·So Professor Murphy, you remember that I
deposed you once before in this matter; correct?
· · · A.· ·Yes, I do.
· · · Q.· ·And that was in November of 2013?
· · · A.· ·Sounds about right, although I couldn't tell
you for sure.
· · · Q.· ·And since that deposition what work have you
done on the Apple case?
· · · A.· ·I have done the work that's reflected in the
declaration that I gave here.· I have continued to read
through the declarations prepared by plaintiffs' experts,
both Professor Noll and Professor Wooldridge, and evaluate
those reports and the claims that are made therein.
· · · Q.· ·Okay.· When you say you've done the work that's
reflected in the declaration, are you referring to the
supplemental report that you submitted together with
Professor Topel dated December 20, 2013?
· · · A.· ·Yes, because that was produced after our
earlier deposition, so when you asked me what we had
worked on since then, that was a component that I
definitely had worked on since that date.

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

recall whether I had primary responsibility for one or the
other.· I know I worked on the whole thing, so I don't
even remember which one I started with.
· · · Q.· ·Did anyone assist you and Professor Topel in
writing the supplemental report?
· · · A.· ·Yes, we had some assistance from Anita Garten
who we work with regularly.· And certainly in preparation
in terms of the tables and other things that went in
there, other people, Ricardo Cossa would be a primary
person who worked on doing a lot of the programming and
statistical work based on our direction that's in the
report.
· · · Q.· ·Anyone else?
· · · A.· ·There would be others.· I don't recall all the
people who worked on the statistical end of things.· In
terms of drafting the report it would be me, Professor
Topel, Anita Garten.· I believe other people looked at it
and gave us their thoughts.· That would be Ricardo Cossa
and Bin Chen I believe would be the other one and possibly
Naraj.
· · · Q.· ·And when you and Professor Topel sat down to
put together the supplemental report, what were you
addressing in that report?
· · · A.· ·We -- I think we state right at the beginning
of the report, we address some of the issues that

Page 239

Page 240

Page 241

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

Professor Noll brought up in his -- I don't know exactly
what the title of his rebuttal or his most recent report
that was filed after our earlier reports.· So we based our
discussion there on the claims made by Professor Noll in
his report.· And we focused on the topics he discussed in
his supplemental report or whatever the specific -- actual
name of it is, not back to his original report, and we
make reference to his original report only when it's
needed for context.· I think this is described directly in
the report that we filed.
· · · Q.· ·And in preparing the supplemental report, did
you or staff at CRA take any additional statistical
analysis?
· · · · · ·MR. KIERNAN:· Object to form.
· · · · · ·THE WITNESS:· Yes, I believe there is
statistical analyses presented in that report.· Yes.· So
we did do statistical analysis for purposes of that
report.
BY MS. SWEENEY:
· · · Q.· ·Can you describe the analyses that you did?
· · · A.· ·They are contained in the report.· It mostly
consisted of variations of the regression model that
Professor Noll estimated, many of the variations we had
done before.· Although since he had changed the
specification in some ways, it was a little different.

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

data that were in the Noll rebuttal report?
· · · A.· ·No, I think we did -- I think we did some
things that were different.· I think the basic thrust of
what we did was very similar.· I believe we did some
different things in this, in this report than we did in
the other.· They are not inconsistent with what we did
before, but we did add some to that.· But the biggest
difference, particularly in terms of the regression
models, would be the basis for those, would have been his
new regression specification rather than the old one.
· · · Q.· ·So what are the different things that you did?
· · · A.· ·You know, I don't recall each one of them
individually so there may be some that I leave out.· For
example, I know we did some F-tests on various blocks of
coefficients in this draft that we had not done in the
earlier one.· We did I think two different versions of
clustering in this particular report that we hadn't -- we
had done just the single version before.· We also examined
the correlation of the residuals somewhat further from
what we had done in the previous example.· So we had done
a few things different.· I don't recall all of them off
the top of my head.
· · · Q.· ·So you said you did F-tests on blocks of
coefficients.· You hadn't done any F-tests prior to this
supplemental report?

Page 242

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

But the basic take of what we did there in terms of
redoing his regression analysis was very similar to what
we had done before.· It just simply worked with his latest
specification rather than the specification that was in
his earlier report.· I think the substance of what we
discussed there is very similar.· We went on and because
he had changed the way he did his weighting and
calculation of his standard errors, we went and did some
new work using the same methodology to illustrate that
there was still a correlation problem with his residuals
from his regression.· That would be different than what we
had in the earlier report, not in the sense that the
methodology is different, is because he had done something
different.· The specifics of the results would be somewhat
different, although the basic conclusion would be the same
that he continues to have a rather severe problem with
correlation between different observations in his dataset
that causes him to greatly understate the standard errors
of his estimates.
· · · Q.· ·So when you say you conducted some new work
using the same methodology that you had used in your
previous reports, and when I say reports here I'm
referring to the earlier Murphy and Topel reports, am I
correct in sort of rephrasing to say that you didn't
conduct any new tests, but you just used the different

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · A.· ·I think -- I don't know if we did some F-tests
or not.· I know we did some specific ones here I believe
than we did before.
· · · Q.· ·And are those reflected in the report?
· · · A.· ·Yes, they are reflected in I believe Table 5.
It's JT-5 or something like that is the table that has
those.
· · · Q.· ·And then you said you did two different
versions of clustering, whereas in your prior report you'd
only done one different version of clustering.· What do
you mean by that?
· · · A.· ·In this one we did I believe clustering by
family as opposed to by family quarter.· I think if I
recall correctly, that's -- I don't want to -- it's either
JT-3 or JT-4.· I can't remember which one, but we did that
in this report.
· · · Q.· ·And did you also do clustering by family
quarter?
· · · A.· ·Yes.· That's a primary thing, but that's what
we had done before, so that wouldn't be different.· That
was how we had done it previously.
· · · Q.· ·And when you say "family," what are you
referring to?
· · · A.· ·It's the iPod family, which is a categorization
that Apple uses to distinguish different models of their

Page 243

Page 244

Page 245

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

iPod, so it might be like an iPod mini, second generation,
given quality level, say good, better or best, one of
those.
· · · Q.· ·And then you said that you also examined
correlation of residuals in a way that you hadn't before.
Can you explain that a little more?
· · · A.· ·I don't recall everything that we put -- that's
in the report, so I just know that we redid that analysis
based on his new results, and I'm not sure everything we
did there is exactly the same as what we did before.· I'd
have to go back and look at the report to tell you.
· · · Q.· ·You also testified that you evaluated the
plaintiffs' experts' declarations, including the
Wooldridge report.· Have you done any subsequent
statistical analysis since reviewing the Wooldridge
report?
· · · A.· ·Yes, I have.· I have looked at the data to
evaluate whether his claims are correct or not, and I
think the data overwhelmingly say that his claims are
incorrect.
· · · Q.· ·And when you say you looked at the data, what
data did you look at?
· · · A.· ·The data based on Professor Noll's regression
data in his regression model.
· · · Q.· ·And do you have -- do you anticipate doing any

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

have considered in preparing his or her report.· And
paragraph 6 of the stipulation order should not be
construed to preclude reasonable questions at deposition
going to the expert's compensation, hours expended in
preparing his or her report and testimony and frequency
and duration of meetings with counsel.
· · · · · ·So what I'm getting at is I think that counsel
for Apple has construed overly broadly the restrictions of
the stipulation.· So let me ask you again -· · · · · ·MR. KIERNAN:· Bonny?
· · · · · ·MS. SWEENEY:· Yeah.
· · · · · ·MR. KIERNAN:· I'm happy for you to ask him
about the categories in 5 or 6, but your question was
whether or not he has been asked to or that he will submit
any additional writing in response to either Drs. Noll's
or Wooldridge which is not covered by Categories 5 or 6.
So you can ask him about 5 or 6.
· · · · · ·MS. SWEENEY:· Let me ask you, David, is there
anything in the stipulation that prohibits oral questions
at deposition about the expert's discussions with the
attorneys?· I see it talks about documents.· I don't see
anything that says I can't ask him what the lawyers asked
him to do.· If you can point that out to me, maybe I'm
missing it, but I don't see it.
· · · · · ·MR. KIERNAN:· Well, I will pull out previous

Page 246

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

further writing of reports or presenting of analyses that
will reflect your conclusions regarding Professor
Wooldridge's analysis?
· · · · · ·MR. KIERNAN:· And I'm going to object and
instruct the witness not to answer any, to the extent it
reveals any communication or requests from counsel.· So
I'm instructing you not to answer the question to the
extent that it would reveal requests from counsel.
BY MS. SWEENEY:
· · · Q.· ·Are you going to answer the question?
· · · A.· ·I have been instructed not to answer, so...
· · · Q.· ·Well, let's break that up a little bit.· So the
stipulation in this case says that certain categories of
data, information, documents and materials need not be
produced, and that includes written correspondence between
an expert and the attorneys, notes taken and prepared by
or for an expert in connection with the matter, including
notes of conversations with the attorneys, but then the
stipulation goes on to say, and what I was paraphrasing
was in paragraph 3 of the stipulation, goes on to say
paragraph 3 -- excuse me, paragraph 5 says, nothing in
paragraph 3 however shall be construed to prevent
substantive deposition questions with respect to
alternative theories, methodologies, variables, data,
production of materials or assumptions that the expert may

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

objections and instructions that you gave Dr. Noll,
relying upon -· · · · · ·MS. SWEENEY:· Good luck with that.· I never
stopped him from answering those kind of questions.
· · · · · ·MR. KIERNAN:· Are you going to continue to
interrupt me?
· · · · · ·MS. SWEENEY:· No.· Go ahead.· Let's create this
record.
· · · · · ·MR. KIERNAN:· Okay.
· · · · · ·-- instructing him not to answer questions
about communications between counsel and himself.· And Xan
Bernay also instructed Dr. Martin not to answer any
questions about communications that she had or that you
had with Dr. Martin -· · · · · ·MS. SWEENEY:· Well -· · · · · ·MR. KIERNAN:· -- based on the stipulation.
· · · · · ·MS. SWEENEY:· Are you done?· I'm sorry. I
don't want to interrupt you.
· · · · · ·MR. KIERNAN:· I am done.
· · · · · ·MS. SWEENEY:· I don't believe that that's a
proper interpretation of the stipulation, and to the
extent you let that pass, I can't explain that.· I'm sure
I never instructed Professor Noll not to answer your
questions on that basis.· And if you can point me to some
deposition testimony, I would be very, very surprised

Page 247

Page 248

Page 249

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

because that is just completely false.· So if you're
instructing the witness not to answer, that is an improper
instruction, and I will take it to the court.· So let me
make sure the record is clear on this.
BY MS. SWEENEY:
· · · Q.· ·Professor Murr- -- strike that.
· · · · · ·Professor Murphy, are you on the advice of
counsel for Apple refusing to answer my question about
communications that you had with Apple's counsel about
work that you have performed or will perform in this
matter?
· · · A.· ·Yeah, I'm taking the advice of the counsel for
Apple.· Since I'm not a lawyer, I can't tell you all the
interpretation, and I can -- best I can do is go with the
advice of counsel.
· · · Q.· ·So you're refusing to answer any question that
I ask about instructions given to you by counsel for
Apple?
· · · A.· ·If that's what I'm instructed to do by counsel
for Apple, that's presumably what I should do.
· · · Q.· ·So if counsel for Apple has instructed you or
provided information to you that you've relied on, are you
going to refuse to answer questions about that based on
the advice of counsel?
· · · · · ·MR. KIERNAN:· Objection.· Object to form.

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

anticipate doing in this matter?
· · · A.· ·I anticipate continuing my work, evaluating the
claims made by Professor Noll and by Professor Wooldridge,
that's ongoing, and I can -- I would, I would presume that
I would continue to do that in the coming days and weeks
and however long it takes.
· · · Q.· ·Are you writing a report to be submitted next
week in opposition to plaintiffs' motion to exclude
portions of your testimony and Professor Topel's
testimony?
· · · A.· ·I don't know whether there is going to be a
report submitted or not.· I -- I know I'm continuing to do
work.· And there is a possibility that that would lead to
a report.· I would understand, but I can't tell you
whether there is going to be one submitted or not.
That's -- I'm just going to do my work and continue to try
to evaluate the claims of plaintiffs' experts.
· · · Q.· ·Did you receive input from Apple's counsel in
your report, your supplemental report?
· · · A.· ·What do you mean by received input from them?
They gave -- they gave us comments.· That's pretty much
standard on drafts of the report.· I don't know.· Usually
that's covered by stipulation, the specifics of those
comments, but again, you guys are the lawyers, so, yes, we
did discuss the report and the contents of the report with

Page 250

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · · · ·THE WITNESS:· I -· · · · · ·MR. KIERNAN:· And that was not the instruction
I gave.
· · · · · ·THE WITNESS:· -- will do my best to answer your
questions.· If there are specific things that I'm not
supposed to answer because they are covered by
stipulation, the best I can do to comply with the desires
of the court I see is to try to follow those directions to
the best I can, and for that I need to rely on counsel.
· · · · · ·MR. KIERNAN:· And just so the record is clear,
I have not made a broad objection as you're stating -that you're suggesting that Dr. Murphy -- this should go
question by question, and if there is a question you want
to ask him about whatever you want to ask him, then we
will go question by question, and if I feel that it's
violating or outside the stipulation, I will assert the
objection and instruct the witness not to answer.· If I
feel that it's not covered by the stipulation, then I
won't assert the objection.
· · · · · ·MS. SWEENEY:· Well, this is an issue on which
we strongly disagree.· I think that we will just see how
it goes, but this is something we will probably have to
raise with the Court.
BY MS. SWEENEY:
· · · Q.· ·Professor Murphy, what additional work do you

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

counsel.
· · · Q.· ·Did you adopt the suggestions of Apple's
counsel in your supplemental report?
· · · A.· ·Not really.
· · · Q.· ·Did they write any of the report, that is,
Apple's counsel?
· · · A.· ·No.
· · · Q.· ·Did you write or did Professor Topel write all
of the supplemental report?
· · · A.· ·No, there would have been help from -- I
believe Anita Garten helped.· Ricardo Cossa maybe did a
little bit.· I know he provided some comments and edits,
but I would say we wrote the report between Bob and
myself.· Some of the cites and footnotes were filled in by
people.· We would say cite to a particular document or
particular book or chapter.· We wouldn't necessarily know
the exact title it would be, you know, cite to Angrist and
Pischke.· Somebody would have to fill it in with the
entire Angrist and Pischke cite.· So if you consider that
writing a report, other people filled in the rest of those
footnotes and things like that.· But the substance was
written by Bob and myself.
· · · Q.· ·How long -- strike that.
· · · · · ·How many hours did you spend preparing the
supplemental report?

Page 251

Page 252

Page 253

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · A.· ·I don't recall.
· · · Q.· ·More than 10?
· · · A.· ·Yeah, it would have been more than 10.
· · · Q.· ·More than 100?
· · · A.· ·It would have been less than 100.
· · · Q.· ·Less than 50?
· · · A.· ·I would assume so.
· · · Q.· ·Less than 40?
· · · A.· ·That I -- now we are getting down to -- my
level of resolution I think is about that level.
· · · Q.· ·Do you know how much time Professor Topel spent
on it?
· · · A.· ·I do not.· You'd have to ask him.
· · · Q.· ·Is there anything today that you would like to
revise or correct in your supplemental report?
· · · A.· ·Nothing that specifically I'm aware of, no.
· · · Q.· ·Is there anything in your prior report that you
want to correct or revise?
· · · A.· ·No, not that I'm aware of.
· · · Q.· ·Have you reviewed the Wooldridge deposition
transcript?
· · · A.· ·I skimmed through it.· I just got it yesterday,
and I was busy at school with things yesterday, so I only
had a chance to skim through it.
· · · Q.· ·And did you review the deposition transcript

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

pricing that he leaves out which biases his results and
causes him to misestimate the impact of any of the iTunes
7.0 release.· He continues to use the wrong but-for world
and the wrong specification of the impact of iTunes 7 on
iPod prices.
· · · · · ·The basic methodology he continues to use is
greatly flawed because of the fact that again his
estimates are not isolating the impact of the challenged
conduct, that simultaneous with the release of iTunes 7
other things in the marketplace changed, including the
models being offered by Apple, the characteristics of
those products, and rendering his analysis sort of invalid
conceptually.· Those are the ones I remember.· There might
be some more conclusions, but those are the basic ones.
· · · Q.· ·And are the bases for all of the opinions that
you've just summarized set forth in your supplemental
report?
· · · A.· ·Yes.· I think we tried to summarize the bases
for those opinions at the time we wrote the report, yes.
· · · Q.· ·Now, you said that Professor Noll has continued
to overstate the significance of his results, and you say
that there is a strong degree of correlation.· Did you
rely upon any economic literature other than the
literature you cited in your first report to support your
conclusions regarding the clustering?

Page 254

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

from Professor Noll's deposition?
· · · A.· ·I did.· I did, but that was a while back.
· · · Q.· ·Can you summarize for me the opinions that are
stated in your supplemental report?
· · · A.· ·Probably not all of them.· I mean, I think
there is some basic ones.· One is I think the first and
foremost is Professor Noll has continued to overstate the
significance of his results by un -- by failing to account
for the strong degree of correlation across transactions,
and that's a very substantial error.· It results in him
calculating standard errors that are roughly 100 times too
small.· I mean, not always exactly that.· Sometimes a lot
more; sometimes less.· But it's a gross misestimate of the
precision of his estimates.
· · · · · ·That the -- his reasons for not taking account
of the correlation are invalid, and we go through each of
those in the report.· The residuals at the family by
quarter level are demonstrably correlated to a high
degree.· The fact that he has all the transactions,
actually nearly -- or nearly all the transactions, what he
calls it's a population rather than a sample doesn't
diminish the need to account for the correlation that's
present in the data that he has.
· · · · · ·That he continues to have omitted variables in
his regression, that important determinants of iPod

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · A.· ·I don't recall what all we cited in our first
report.· There is certainly a broad economic literature
that supports what we put forward there.· I would say what
we cite here is certainly you could find other cites that
would echo the same points.· The articles we have cited,
you know, sort of Cameron and Miller and Donald and Lang
and Angrist and Pischke and the Hansen, Chris Hansen's
work, I mean a number of things that we have referred to
and looked at all support the same thing.
· · · · · ·If you want to see more, I think Cameron and
Miller have a pretty extensive discussion of the
underlying further literature that you might want to look
at if you wanted more literature, but we didn't cite those
directly.
· · · Q.· ·Do you consider yourself an expert on
clustering issues?
· · · A.· ·I would consider myself an expert at applied
econometrics generally and in particular methods to be
used for these type of data.· I have spent my career
analyzing aggregate -- more aggregated phenomena using
microdata, so this is an area I'm very familiar with.· So
it's something that I would say, yeah, I would consider
myself very well versed in.
· · · Q.· ·Have you ever written any articles that have
been published specifically relating to clustering?

Page 255

Page 256

Page 257

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · A.· ·I don't think I published any specific articles
on clustering.· I have certainly talked about how you use
data of this type where you have market-level phenomena
you're interested in and micro-level data used to estimate
it.
· · · · · ·I did a lot of well-cited work on the wage
structure, which we made exactly the point that we are
talking about here, that even though you might have
hundreds of thousands of observations from underlying
datasets, when it comes to estimating the determinants of
market prices, you have much less data because in fact you
only have data on a limited number of market equilibrium.
And that's exactly the same point that's being made here.
And my work in that area goes back probably almost 30
years.
· · · Q.· ·Now, why don't we go ahead and mark the
supplemental report as Murphy Exhibit 6, because there
were five exhibits marked in your prior deposition.
· · · · · ·(Exhibit 6 marked.)
· · · Q.· ·Now, you said that you and Professor Topel
jointly wrote the report; correct?
· · · A.· ·That's correct.
· · · Q.· ·For those parts that you didn't yourself write,
do you agree with them?
· · · A.· ·Yeah, I mean, I worked on every part of the

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · Q.· ·Professor Noll disagrees with you that there is
a clustering problem; isn't that correct?
· · · A.· ·Yes, but he disagrees by just denying it, not
by presenting evidence that there is no correlation.· He
doesn't -- he doesn't calculate the correlation and say,
look, the correlation is low.· He never -- he never does
that.· I mean, we present a graphical representation of
just how highly correlated the residuals are and uses -you know, he just asserts things without really doing any
analysis.
· · · Q.· ·That's interesting.
· · · · · ·Did you perform any cluster analyses that are
not presented in your exhibits?
· · · A.· ·What does that mean?
· · · Q.· ·Well, you and your staff at CRA performed some
analyses on the clusters; correct?
· · · A.· ·We have continued to work on the clustering
analysis.· Well, it's actually not really the clustering
analysis.· Analysis of Professor Noll's model and analysis
of the error terms in that model and their properties.· So
we have continued to work on that subsequent to the
report.· Is that what you're asking?
· · · Q.· ·No.· What I want to know is, did you perform
any analyses that relate to your opinion regarding
clustering that are not presented in the tables in your

Page 258

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·

report.· I mean, we didn't like write them and then glue
them together and send them in.· We worked on them back
and forth.· We talked about it before we ever wrote
anything.· Even if one of us wrote the paragraph the first
time, the other one went through it, and we ended up with
something we were both satisfied with.
· · · Q.· ·So you agree with all the statements in the
report; is that correct?
· · · A.· ·To the best of my knowledge, yes.
· · · Q.· ·In paragraph 5.a., you say that Professor Noll
simply ignores the clustering problem.· In your opinion
did Professor (sic) in his rebuttal -· · · · · ·MR. KIERNAN:· Page 2.
BY MS. SWEENEY:
· · · Q.· ·-- report ignore the clustering problem?

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

supplemental report?
· · · A.· ·Are you referring to analyses done before the
report or after the report?
· · · Q.· ·Let's start with before the report.
· · · A.· ·Yeah, I'm sure we did additional analyses that
we are not relying on.· I don't -- there is nothing that
would be inconsistent with what we have done there.· There
are often many ways to do things, and when it comes time
to write a report, you settle on one that you think
accurately what you did.· But nothing that I'm relying on,
no.
· · · Q.· ·And is that true also for your initial report,
that is, that you conducted certain analyses relating to
your clustering opinions that are not reflected in the
tables and exhibits attached to your report?
· · · · · ·MR. KIERNAN:· Objection; argumentative.
· · · · · ·THE WITNESS:· Nothing that I would be relying
on.
BY MS. SWEENEY:
· · · Q.· ·But nonetheless, is it your testimony, let's
start with the supplemental report, that those analyses
that are not reflected in your exhibits in which you're
not relying on are nonetheless consistent with your
opinions?
· · · A.· ·Yes, I would say the analysis we have done is

Page 259

Page 260

65

·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

And we settled on quarter
because we thought it represented a reasonable compromise
between taking account of that correlation while still
allowing for some independence over time.· I think that's
the best justification you have for focusing on quarters.
If somebody wanted to aggregate further, I think that
could be reasonable.· I think aggregating less is going in
the wrong direction.· I think I would continue to stand by
that.· And certainly doing what Professor Noll did because
it neither aggregates over time nor across transactions at
a point in time doesn't make any sense.
BY MS. SWEENEY:
· · · Q.· ·So you think it's going in the wrong direction
to cluster at any shorter time interval, but you don't
know in fact whether your staff at CRA conducted the
clustering analysis by week; correct?
· · · · · ·MR. KIERNAN:· Object to form.
· · · · · ·THE WITNESS:· I don't believe they did because
I didn't ask them to do that because I wouldn't have
thought going in that direction made much sense.· It's not
something that would tell -- economics or the data or the

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

allow for clustering.
· · · · · ·Or when people analyze -- go back to the
original work of Fama and McBeth.· They have -- you know,
they have all the stock market data.· They have all the
closing prices on a daily basis for all the stocks in
their universe.· They don't have a sample of dates.· They
got every date, and they have got, you know, all the
stocks in their universe.· They didn't draw -- they didn't
draw a random sample.· They had that.· And Fama and
McBeth, they essentially did original work actually kind
of related to clustering.· They used a different
methodology, but they were concerned with exactly that
same phenomena.
· · · Q.· ·Anything else?
· · · A.· ·I mean, there is bunches of them, but there is
nothing, there is nothing unusual about that, about doing
it.
· · · · · ·I think Professor Wooldridge in his textbook
talks about the Michigan teachers and clustering in the
context of -· · · · · ·(A brief interruption.)
· · · A.· ·So he has one on Michigan teachers I believe in
his textbook where again he is dealing essentially with
population data.· So yeah, it's not hard to find examples
of that.

Page 266

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

documents or anything in this case would tell me to move
in that direction.· So that's not something I would have
instructed them to do, at least as I recall.· I can't tell
you what all -- anything they did, but not that I know of.
It's the best answer I can give.
BY MS. SWEENEY:
· · · Q.· ·You testified a little bit ago about the
difference that Professor Noll asserted between samples of
data and the population of data.· Are you familiar with
any other instance in which the kind of adjustment that
you performed with respect to the standard errors was used
to adjust standard errors for clustering when the dataset
was a population rather than a sample?
· · · A.· ·Yeah.· I mean, for example, in Professor
Wooldridge's work that he did on unilateral divorce, the
sample -- the data on divorce there are population.· They
are the divorces that occurred in each state from vital
statistics, which is a population, not a, not a sample.
The -· · · Q.· ·Anything else?
· · · A.· ·Yeah.· I mean, in like price-fixing cases and
things like that where you have all the transactions for a
company, you don't -- and you're doing a price analysis,
you typically don't treat all the prices as independent
because they are determined by common factors.· You would

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · Q.· ·So I asked you whether you tested weekly -- I
asked about weekly.· Did I ask about monthly clusters?
Did you conduct that analysis?
· · · A.· ·You know, I don't recall whether we did or not.
It's not something I remember.· We might have done that as
a sensitivity check back when.· It's not something I have
done recently.· Might have done that before the first
report, but I don't recall one way or the other.
· · · Q.· ·But if you -· · · A.· ·It's not relying on it for sure.
· · · Q.· ·If you did do that, it's not reflected in the
report or its exhibits or attachments; correct?
· · · A.· ·Yeah, because I wasn't relying on that. I
don't recall doing it specifically, but I don't want to
say something that's not true, so best I can say is I
don't recall.
· · · Q.· ·So you don't know whether if you created
clusters at the weekly level, for example, the results
would be statistically significant?
· · · A.· ·Yeah, but -· · · · · ·MR. KIERNAN:· Object to form.
· · · · · ·THE WITNESS:· But if you do that and find the
standard errors fall significantly, it's mostly because of
the fact that you have inappropriately ignored correlation
that exists in the data.· So that's what I'm saying.· It's

Page 267

Page 268

Page 277

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

that if you wanted to.· You could do lots of different
things.· But the whole point is his analysis is not valid.
He assumes that they are independent regardless of which
one you did.
BY MS. SWEENEY:
· · · Q.· ·If you have a situation where there are no
clusters and you do this kind of clustering analysis, is
that harmless error in your view?
· · · A.· ·You know, I'm going to say, I think it's, I
think it's -- if you suspect there is correlation, which
we have strong reason to do in this case, economics tells
us there will be correlation, the data tells us there is
correlation, the documents in this case tells us there is
going to be correlation, there is just so overwhelming
prior that there will be correlation, then I think it's
proper to allow for that correlation.· I think that is the
only prudent approach.· To just bury your head in the sand
and ignore the fact that they are correlated as Professor
Noll does doesn't make any sense.
· · · Q.· ·I'm going to move to strike as nonresponsive,
and I'm going to ask the court reporter to read my
question back.
· · · · · ·MR. KIERNAN:· I oppose the motion.
· · · · · ·(Record read.)
· · · · · ·THE WITNESS:· I don't want to -- here is what I

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

say there is very little downside to allowing for that
clustering, and the reason is it doesn't create a bias.
It doesn't bias your standard errors upward.· If anything,
it's going to tend to make your standard errors a little
bit smaller is what the literature suggests.· It's not
generally going to lead to an upward bias in your standard
errors.
BY MS. SWEENEY:
· · · Q.· ·Now, you added a number of characteristics to
Professor Noll's regression; correct?
· · · A.· ·Yes, this -- that's not new to this report.· We
had done that in the previous report as well.
· · · Q.· ·And are the -- where in your report are those
characteristics listed?
· · · A.· ·Well, they are listed in the regression tables
themselves.
· · · Q.· ·Okay.· Point to me a particular table that -· · · A.· ·Like just go to -- well, might as well just
start at the beginning.· Go to Table 1.· They are listed
at the bottom of the table.· Well, not the very bottom,
but the bottom of the coefficient portion.· See where
there is like a white space in the first three columns and
then there is numbers in the last two columns?· Those
would be the additional characteristics.
· · · Q.· ·So the first one is HP_OEM?

Page 278

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

wanted to say is, it's -- the right way to proceed is to
allow for the correlation if you think it's going to be
there.· You wouldn't want to allow for correlation if you
had no reason to believe it were there.· That doesn't make
a whole lot of sense, but when economics tells you there
is going to be correlation, when the facts of the case
tells you there is going to be correlation, you should
allow for that correlation.
· · · · · ·Now, is it harmless?· I should say it's the
opposite of harmless.· It's necessary; it's needed.· And
there is no reason to believe there isn't correlation in
this case.· You would expect that there would be
correlation.
BY MS. SWEENEY:
· · · Q.· ·Okay.· Well, listen to my question.· If you
have a case where there is no reason to believe there is
correlation and you conduct this kind of clustering, is
that harmless error in your view?
· · · · · ·MR. KIERNAN:· Objection; asked and answered.
· · · · · ·THE WITNESS:· I would say it will not -- if we
do it -- I don't want to apply it to an abstract world
other than this one.· In the context like this given the
number of clusters that we have, which is upwards of 300
in one case and upwards of 400 in the other case, and
given the nature of the data that we have here, I would

·1·
·2·
·3·
·4·
·5·
·6·
·7·
·8·
·9·
10·
11·
12·
13·
14·
15·
16·
17·
18·
19·
20·
21·
22·
23·
24·
25·

· · · A.· ·Correct.· And it runs down to
log_recharge_hours.
· · · Q.· ·What is HP_OEM?
· · · A.· ·That was the specific iPods that were I believe
HP-branded at one point in time, so they were a little bit
different.· So that allows for it was an OEM iPod as
opposed to an Apple iPod.· That sort of refers to some
specific models.
· · · Q.· ·What about -- what is USB?
· · · A.· ·USB is whether the Ap- -- whether the iPod is
USB-compatible.· Some of the early generation iPods were
not USB-compatible.· They were FireWire based.
· · · Q.· ·And so the next one, FireWire, is that sort of
the opposite then of USB?· Either it has USB or it has
FireWire?
· · · A.· ·No.· There was some that I think had both,
which is why there -- both effects can be there.· The
majority either had one or the other as I recall, and I
don't remember the fraction that had both.· There were
some particular models as I recall that had both.· That's
why they are both in there.
· · · Q.· ·So do you know whether all iPods sold to class
members during the class period had USB capability?
· · · A.· ·I'd have to go back and check.· I don't recall.
But that doesn't affect whether you'd want a control for

Page 279

Page 280

Page 345

Page 346

·1· ·(Concluded at 12:07 p.m.)

·1· · · · · · ·DECLARATION UNDER PENALTY OF PERJURY

·2

·2· Case Name:· The Apple iPod iTunes Anti-Trust Litigation

·3

·3· Date of Deposition:· 1/8/2014

·4

·4· Job No:· 10009199

·5

·5

·6

·6· · · · · I, KEVIN M. MURPHY, Ph.D., the witness herein,

·7

·7· declare under penalty of perjury that I have read the
·8· foregoing in its entirety; and that the testimony

·8

·9· contained therein, as corrected by me, is a true and

·9

10· accurate transcription of my testimony elicited at said

10

11· time and place.

11

12· · · · · Executed this ____ day of ___________________,

12

13· 2014, at _____________________________.

13

14

14

15· _________________________· · · · · · · · ______________

15

· · KEVIN M. MURPHY, Ph.D.· · · · · · · · · · · ·Date

16

16

17

17

18

18

19

19

20

20

21

21

22

22

23

23

24

24

25

25

Page 231

Page 348

·1· · · · · · · · ·DEPOSITION ERRATA SHEET

·1· STATE OF ARIZONA· · · ·)

·2· Page No. ____ Line No. ____

· · · · · · · · · · · · · ·)· ss.

·3· Change: ________________________________________________

·2· COUNTY OF MARICOPA· · ·)

·4· Reason for change: _____________________________________

·3· · · · · · ·BE IT KNOWN that the foregoing deposition was

·5· Page No. ____ Line No. ____

·4· taken before me, Cathy A. Miccolis, RPR, a Certified

·6· Change: ________________________________________________

·5· Reporter, Certificate #50068, for the State of Arizona,

·7· Reason for change: _____________________________________

·6· and by virtue thereof authorized to administer an oath;

·8· Page No. ____ Line No. ____

·7· that the witness before testifying was duly sworn by me to

·9· Change: ________________________________________________
10· Reason for change: _____________________________________
11· Page No. ____ Line No. ____
12· Change: ________________________________________________
13· Reason for change: _____________________________________
14· Page No. ____ Line No. ____
15· Change: ________________________________________________
16· Reason for change: _____________________________________
17· Page No. ____ Line No. ____
18· Change: ________________________________________________
19· Reason for change: _____________________________________

·8· testify to the whole truth; that the questions propounded
·9· to the witness and the answers of the witness thereto were
10· taken down by me in shorthand and thereafter reduced to
11· print by computer-aided transcription under my direction;
12· that pursuant to request, notification was provided that
13· the deposition is available for review and signature; that
14· the transcript consisting of pages 233 through 348 is a
15· full, true and accurate transcript of all proceedings and
16· testimony had and adduced upon the taking of said
17· deposition, all done to the best of my skill and ability.
18· · · · · I FURTHER CERTIFY that I am in no way related to
19· nor employed by any of the parties hereto nor am I in any

20· Page No. ____ Line No. ____

20· way interested in the outcome hereof.

21· Change: ________________________________________________

21· · · · · DATED at Phoenix, Arizona, January 9, 2014.

22· Reason for change: _____________________________________

22

23

23· · · · · · · · · · · · · · _____________________________

24· _________________________· · · · · _____________________

· · · · · · · · · · · · · · · Cathy A. Miccolis, RPR, CRR

· · KEVIN M. MURPHY, PH.D.· · · · · · · · · · Dated

24· · · · · · · · · · · · · · Certified Reporter #50068

25

25