Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 1 of 288
` 919
`
` 1
`
` 2
`
` 3
`
` 4
`
` 5
`
` 6
`
` 7
`
` 8
`
` 9
`
`10
`
`11
`
`12
`
`13
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE DISTRICT OF COLORADO
`
`
`Civil Action No. 21-cr-00229-RBJ
`
`UNITED STATES OF AMERICA,
`
` Plaintiff,
`
`vs.
`
`1. DAVITA INC., and
`2. KENT THIRY,
`
` Defendants.
`_______________________________________________________________
`
`
`
`REPORTER'S TRANSCRIPT
`TRIAL TO JURY - DAY SIX
`
`
`_______________________________________________________________
`
`
`Proceedings before the HONORABLE R. BROOKE JACKSON,
`
`14
`
`Senior Judge, United States District Court for the District of
`
`15
`
`Colorado, continuing at 8:32 a.m., on the 11th day of April,
`
`16
`
`2022, in Courtroom A902, United States Courthouse, Denver,
`
`17
`
`Colorado.
`
`
`
`
`
`
`
`
`
`
`
`
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`THERESE LINDBLOM, Official Reporter
`901 19th Street, Denver, Colorado 80294
`Proceedings Reported by Mechanical Stenography
`Transcription Produced via Computer
`
`
` 919
`
` 1
`
` 2
`
` 3
`
` 4
`
` 5
`
` 6
`
` 7
`
` 8
`
` 9
`
`10
`
`11
`
`12
`
`13
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE DISTRICT OF COLORADO
`
`
`Civil Action No. 21-cr-00229-RBJ
`
`UNITED STATES OF AMERICA,
`
` Plaintiff,
`
`vs.
`
`1. DAVITA INC., and
`2. KENT THIRY,
`
` Defendants.
`_______________________________________________________________
`
`
`
`REPORTER'S TRANSCRIPT
`TRIAL TO JURY - DAY SIX
`
`
`_______________________________________________________________
`
`
`Proceedings before the HONORABLE R. BROOKE JACKSON,
`
`14
`
`Senior Judge, United States District Court for the District of
`
`15
`
`Colorado, continuing at 8:32 a.m., on the 11th day of April,
`
`16
`
`2022, in Courtroom A902, United States Courthouse, Denver,
`
`17
`
`Colorado.
`
`
`
`
`
`
`
`
`
`
`
`
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`THERESE LINDBLOM, Official Reporter
`901 19th Street, Denver, Colorado 80294
`Proceedings Reported by Mechanical Stenography
`Transcription Produced via Computer
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 2 of 288
` 920
`
` 1
`
` 2
`
` 3
`
` 4
`
` 5
`
` 6
`
` 7
`
` 8
`
` 9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`A P P E A R A N C E S
`
`MEGAN S. LEWIS, WILLIAM JEFFERSON VIGEN, SARA MICHELLE
`CLINGAN and ANTHONY WILLIAM MARIANO, Attorneys at Law, U.S.
`Department of Justice, Antitrust Division, Washington Criminal
`II Section, 450 5th Street N.W., Washington, DC, 20530,
`appearing for the Government.
`
`
`JOHN C. DODDS, Attorney at Law, Morgan Lewis & Bockius
`LLP, 1701 Market Street, Philadelphia, Pennsylvania, 19103,
`appearing for Defendant DaVita.
`
`
`JOHN CLAYTON EVERETT, JR., Attorney at Law, Morgan
`Lewis & Bockius LLP, 1111 Pennsylvania Avenue NW, Washington,
`DC, 20004, appearing for Defendant DaVita.
`
`
`JOHN F. WALSH, III, Attorney at Law, WilmerHale LLP,
`1225 17th Street, Suite 2600, Denver, Colorado, 80220,
`appearing for Defendant DaVita.
`
`
`THOMAS MELSHEIMER, Attorney at Law, Winston & Strawn
`LLP, 2121 North Pearl Street, 9th Floor, Dallas, Texas, 75201,
`appearing for Defendant Thiry.
`
`
`JUANITA ROSE BROOKS, Attorney at Law, Fish &
`Richardson, 12860 El Camino Real, Suite 400, San Diego,
`California, 92130, appearing for Defendant Thiry.
`
`* * * * *
`
`P R O C E E D I N G S
`
`(In open court at 8:32 a.m.)
`
`THE COURT: Good morning. How do you wish to proceed?
`
`MR. VIGEN: Your Honor, we talked to defense counsel,
`
`20
`
`and we believe it makes sense to go right to cross-examination
`
`21
`
`of the expert as part of the Daubert hearing.
`
`22
`
`23
`
`24
`
`THE COURT: All right.
`
`Good morning.
`
`MR. VIGEN: Before I begin, I just for the record,
`
`25
`
`want to give a thanks to Terri Lindblom for her great work last
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 3 of 288
` 921
`
`Pierre Cremieux - Cross
`
` 1
`
`week and working over the weekend to get us some transcripts.
`
` 2
`
`We appreciate that, Terri.
`
` 3
`
` 4
`
`(PIERRE CREMIEUX, DEFENDANTS' WITNESS, SWORN)
`
`CROSS-EXAMINATION
`
` 5
`
`BY MR. VIGEN:
`
` 6
`
`Q. Good morning, Dr. Cremieux.
`
` 7
`
`A. Good morning.
`
` 8
`
`Q. My name is William Vigen. I'd like to discuss today your
`
` 9
`
`opinions that you filed in your expert report in this matter in
`
`10
`
`Sections 5 and 6. And just briefly, that's where you compared
`
`11
`
`DaVita turnover and compensation data against turnover and
`
`12
`
`compensation data from industry benchmarks that you selected.
`
`13
`
`Is that generally accurate?
`
`14
`
`A. Yes. I compared the movement in the turnover and
`
`15
`
`compensation for DaVita with the movement in compensation -- in
`
`16
`
`turnover and compensation for a benchmark, which was the
`
`17
`
`healthcare industry.
`
`18
`
`Q. Okay. And that type of analysis -- sorry. That type of
`
`19
`
`analysis is called a difference-in-difference approach by
`
`20
`
`economists; correct?
`
`21
`
`A. That's right.
`
`22
`
`Q. Okay. For a valid difference-in-difference analysis, it's
`
`23
`
`important that you're comparing apples to apples; would you
`
`24
`
`agree with me on that?
`
`25
`
`A. It's not a technical term; but generally speaking, I
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 4 of 288
` 922
`
`Pierre Cremieux - Cross
`
` 1
`
`understand the spirit.
`
` 2
`
`Q. Yeah. So the one apple is the DaVita data, and the other
`
` 3
`
`apple is supposed to be your benchmark data. Is that right?
`
` 4
`
`A. Not quite. One apple, if you will, is the change in DaVita
`
` 5
`
`compensation; and the other one is the change in the benchmark.
`
` 6
`
`Q. And if you're not comparing apples to apples, essentially,
`
` 7
`
`you haven't done a valid analysis; would you agree with that?
`
` 8
`
`A. Again, it's very vague. Apples to apples, I don't know
`
` 9
`
`what that means. But generally speaking, if you didn't do it
`
`10
`
`right, then it's wrong.
`
`11
`
`Q. Okay. So let's talk about how you're supposed to do it
`
`12
`
`right. The assumption is, if you're comparing or using an
`
`13
`
`appropriate benchmark -- what I'm referring to is apples to
`
`14
`
`apples, but an appropriate benchmark -- that the various
`
`15
`
`economic factors and variables that could affect the turnover
`
`16
`
`and compensation data are essentially impacting the DaVita data
`
`17
`
`and the benchmark data you selected in relatively similar ways
`
`18
`
`overall. Is that the idea?
`
`19
`
`A. That's fair. Yeah.
`
`20
`
`Q. Okay. So one variable that could impact turnover or
`
`21
`
`compensation is how fast the economy was growing; right?
`
`22
`
`A. Yes. That's a good example.
`
`23
`
`Q. Another variable could be overall inflation; would you
`
`24
`
`agree?
`
`25
`
`A. Yes. That's right, too.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 5 of 288
` 923
`
`Pierre Cremieux - Cross
`
` 1
`
`Q. One other could be wage growth?
`
` 2
`
`A. Could be, sorry?
`
` 3
`
`Q. Wage growth.
`
` 4
`
`A. I don't think that's quite right. It depends on which wage
`
` 5
`
`growth you're looking at. Right? If you're looking at
`
` 6
`
`compensation, for example, which I am, then since what I'm
`
` 7
`
`looking at is whether there is a difference between the two in
`
` 8
`
`the rate at which they grow, that's actually the variable I'll
`
` 9
`
`be interested in, as opposed to one that I would control for,
`
`10
`
`for example.
`
`11
`
`Q. Okay. So another variable could be job satisfaction?
`
`12
`
`A. Okay. So now we have to -- you mean specifically in this
`
`13
`
`application; right?
`
`14
`
`Q. Yes, sir.
`
`15
`
`A. No. I don't think that's right either, because I'm not
`
`16
`
`trying to make a statement about the general economy on one
`
`17
`
`hand and DaVita on the other. The statement I'm making is
`
`18
`
`different from that, and it's important for that point. The
`
`19
`
`statement I'm making is that, as I look at the change in the
`
`20
`
`rate of growth or shrinkage of compensation or turnover, I want
`
`21
`
`to make sure that I anchor it to something so that what I'm
`
`22
`
`looking at is real, if you will. And what I'm anchoring it is
`
`23
`
`the change at the benchmark. And so if turnover is higher,
`
`24
`
`say, at -- or if -- sorry -- if job satisfaction is higher at,
`
`25
`
`say, DaVita relative to the benchmark, that's okay. Because
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 6 of 288
` 924
`
`Pierre Cremieux - Cross
`
` 1
`
`since what I'm interested in is the change, unless there is
`
` 2
`
`something which would lead somehow the economy to have a change
`
` 3
`
`in job satisfaction, which is not reflected at DaVita because
`
` 4
`
`somehow it's insulated from the economy -- which I don't
`
` 5
`
`believe it is -- then my comparison is going to be valid.
`
` 6
`
`In other words, if you have my charts in mind, the
`
` 7
`
`point I'm making is that I think it was -- you know, there was
`
` 8
`
`a chart with 13 percent versus 9 percent. That difference is
`
` 9
`
`not one that I'm particularly interested in, but it's the one
`
`10
`
`you're referring to. What I'm interested is how the change
`
`11
`
`compares between the two groups.
`
`12
`
`Q. Okay. So I'd like to draw out a little bit of what you
`
`13
`
`just said. So you said if the change in job satisfaction is
`
`14
`
`different over time, then your analysis is not valid.
`
`15
`
`If the change in job satisfaction at DaVita changes
`
`16
`
`over time relative to the national benchmark, your analysis is
`
`17
`
`not valid?
`
`18
`
`A. What I'm saying that if there was a significant -- if there
`
`19
`
`was a reason --
`
`20
`
`Q. Sir --
`
`21
`
`A. Sorry.
`
`22
`
`Q. My question is --
`
`23
`
`THE COURT: Are you under the impression you're
`
`24
`
`communicating with the Court, or are you just having a nice
`
`25
`
`discussion with one economist to another? Because I don't know
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 7 of 288
` 925
`
`Pierre Cremieux - Cross
`
` 1
`
`what you're talking about.
`
` 2
`
`MR. VIGEN: Okay. Let me back up.
`
` 3
`
`BY MR. VIGEN:
`
` 4
`
`Q. You agree, Dr. Cremieux, that, essentially, various
`
` 5
`
`variables can impact and cause turnover and compensation data
`
` 6
`
`to change; correct?
`
` 7
`
`A. Yes.
`
` 8
`
`Q. Okay. And we talked about some variables that you
`
` 9
`
`acknowledged could impact that change. You at least
`
`10
`
`acknowledged overall inflation or the economy growing; right?
`
`11
`
`A. In the case of compensation, inflation would matter. I
`
`12
`
`think in the case of job turnover, you took the example of job
`
`13
`
`satisfaction, so that may matter.
`
`14
`
`Q. Okay. So one of the variables that might matter is job
`
`15
`
`satisfaction at DaVita?
`
`16
`
`A. We're now back on the job turnover; right?
`
`17
`
`Q. I'm just asking if job satisfaction as an economic variable
`
`18
`
`could impact turnover and compensation numbers at DaVita.
`
`19
`
`A. It's possible.
`
`20
`
`Q. Okay. The assumption is, using a difference-in-difference
`
`21
`
`approach, that job satisfaction at DaVita relative to the
`
`22
`
`benchmarks you chose is roughly the same, that they're
`
`23
`
`impacting those numbers in relatively the same way?
`
`24
`
`A. No, it's not.
`
`25
`
`Q. Okay. Can you explain why not?
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 8 of 288
` 926
`
`Pierre Cremieux - Cross
`
` 1
`
`A. Because in that analysis, what I'm doing is I'm looking at
`
` 2
`
`the trend, the growth, if you will, or reduction in
`
` 3
`
`compensation for DaVita. And all I'm doing with the benchmark
`
` 4
`
`is I want to make sure that if I observe a growth in
`
` 5
`
`compensation at DaVita, I know how to interpret it, given
`
` 6
`
`that -- if, for example, in the benchmark, the growth was twice
`
` 7
`
`as fast, it means that, really, DaVita is lagging. If the
`
` 8
`
`growth is half as fast, it means DaVita is going faster. So
`
` 9
`
`what I'm doing here is really just using the benchmark to help
`
`10
`
`interpret the DaVita data. Not making any statements about
`
`11
`
`whether the fact that I observed a 13 percent growth in
`
`12
`
`compensation at DaVita and a 9 percent growth in compensation
`
`13
`
`for the benchmark means anything. Doesn't make any difference.
`
`14
`
`Q. Okay. What I would like to focus on is how to do a proper
`
`15
`
`difference-in-difference analysis, not what you did.
`
`16
`
`A. Okay.
`
`17
`
`Q. Okay. So let's just back up to the first principles.
`
`18
`
`A. Okay.
`
`19
`
`THE COURT: So what you're saying is you don't want to
`
`20
`
`focus on what he did; you want to focus on some analysis of
`
`21
`
`yours?
`
`22
`
`MR. VIGEN: No. I'm focusing on the method that he
`
`23
`
`chose to use, which is the difference-in-difference method.
`
`24
`
`THE COURT: Let's see if I can even understand what
`
`25
`
`we're talking about.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 9 of 288
` 927
`
`Pierre Cremieux - Cross
`
` 1
`
`Sir, in section 5, as I read it -- and probably
`
` 2
`
`putting it more simply than you do -- you're looking at
`
` 3
`
`DaVita's turnover rate for, during, and after the conspiracy
`
` 4
`
`periods, and you're looking at a national -- you call it a
`
` 5
`
`benchmark -- average turnover rate for senior executives; yes?
`
` 6
`
` 7
`
`THE WITNESS: That's exactly right. Yes.
`
`THE COURT: And you're comparing DaVita's turnover
`
` 8
`
`rate to the benchmark during this period of time; yes?
`
` 9
`
`THE WITNESS: I'm comparing the two changes. So they
`
`10
`
`go up, both at different rates, and I'm just looking at the
`
`11
`
`changes of the two.
`
`12
`
`THE COURT: You say that during the period starting
`
`13
`
`before our conspiracy periods and continuing all the way
`
`14
`
`through and even after the conspiracy periods, the national
`
`15
`
`turnover rate for executives was 8 to 9 percent.
`
`16
`
`17
`
`THE WITNESS: That's right.
`
`THE COURT: And you said that during that same period,
`
`18
`
`DaVita's rate of turnover started at 4 percent and grew up to 7
`
`19
`
`to 8 percent.
`
`20
`
`THE WITNESS: I don't remember the numbers off the top
`
`21
`
`of my head; but, yes, I make statements about --
`
`22
`
`THE COURT: I think those are the numbers, because I
`
`23
`
`just read them over the weekend.
`
`24
`
`And then you say, thus, DaVita's rate of turnover
`
`25
`
`increased relative to the benchmark. And you say that wouldn't
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 10 of 288
` 928
`
`Pierre Cremieux - Cross
`
` 1
`
`happen, in your opinion, if there was a labor market allocation
`
` 2
`
`agreement in place.
`
` 3
`
` 4
`
` 5
`
`THE WITNESS: That's exactly right. Yes.
`
`THE COURT: That's section 5.
`
`Section 6, you're, in theory, comparing DaVita's
`
` 6
`
`salaries during, before, and after the conspiracy period to the
`
` 7
`
`national average; yes?
`
` 8
`
` 9
`
`THE WITNESS: Yes.
`
`THE COURT: And you say DaVita's median annualized
`
`10
`
`salaries for senior-level executives exceeded the national
`
`11
`
`benchmark by 50 percent before the conspiracy period. In other
`
`12
`
`words, you worked at DaVita, you got paid better than the
`
`13
`
`national average.
`
`14
`
`15
`
`THE WITNESS: That's right.
`
`THE COURT: During the conspiracy period, it wasn't
`
`16
`
`50 percent anymore; it was 52 percent higher.
`
`17
`
`18
`
`THE WITNESS: That's right.
`
`THE COURT: After the conspiracy period, you say it
`
`19
`
`was now only 45 percent higher. So you say the highest DaVita
`
`20
`
`average salary compared to the benchmark occurred during the
`
`21
`
`conspiracy period. And then from that, you concluded if there
`
`22
`
`was an allocation agreement in effect, that shouldn't have
`
`23
`
`happened.
`
`24
`
`THE WITNESS: Correct. I would expect the opposite.
`
`25
`
`Yes.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 11 of 288
` 929
`
`Pierre Cremieux - Cross
`
` 1
`
` 2
`
` 3
`
`THE COURT: That's what section 5 and 6 are about.
`
`THE WITNESS: That's right.
`
`THE COURT: Okay. Now, what's your problem with it,
`
` 4
`
`that he's so full of beans that his opinion can't even be
`
` 5
`
`heard?
`
` 6
`
`BY MR. VIGEN:
`
` 7
`
`Q. So, Dr. Cremieux, can you please explain to the Court what
`
` 8
`
`the parallel trends assumption or the common trends assumption
`
` 9
`
`is with respect to the difference-in-difference approach.
`
`10
`
`A. So the assumption here is that there is validity in looking
`
`11
`
`at how the DaVita compensation increased and comparing it with
`
`12
`
`the change in compensation in a sector which DaVita used as its
`
`13
`
`benchmark for its compensation in its own document, which is
`
`14
`
`the outpatient healthcare sector. So I'm not doing anything
`
`15
`
`here that is unusual, that's methodologically problematic or
`
`16
`
`inconsistent with very much the way in which DaVita itself was
`
`17
`
`thinking about what benchmark it should use when it's thinking
`
`18
`
`about its own employees and how to compensate them.
`
`19
`
`MR. VIGEN: Could we please show the witness
`
`20
`
`Government Exhibit 588. And if we can go to page 3.
`
`21
`
`BY MR. VIGEN:
`
`22
`
`Q. Sir, I was trying to get you to explain in laymen's terms
`
`23
`
`for the Court the parallel trends assumption; but I'd like to
`
`24
`
`read you this explanation of it.
`
`25
`
`"The parallel trend assumption is the most critical of
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 12 of 288
` 930
`
`Pierre Cremieux - Cross
`
` 1
`
`the above four assumptions to ensure internal validity of
`
` 2
`
`difference-in-difference models" --
`
` 3
`
` 4
`
`THE COURT: What are you reading from?
`
`MR. VIGEN: This is an article explaining the
`
` 5
`
`difference-in-difference analysis on the Columbia University
`
` 6
`
`website. I'm trying to see if the witness agrees with this to
`
` 7
`
`try to get this in simpler terms.
`
` 8
`
`THE COURT: An article on the Columbia University
`
` 9
`
`website?
`
`10
`
`MR. VIGEN: Describing the difference-in-difference
`
`11
`
`approach, and I'd like the witness to see if he agrees with
`
`12
`
`this.
`
`13
`
`14
`
`15
`
`THE COURT: How long is the article?
`
`MR. VIGEN: Four pages.
`
`THE COURT: Do you want to take the time to read it,
`
`16
`
`or are you familiar with it already?
`
`17
`
`THE WITNESS: I'd love to have it in front of me just
`
`18
`
`so I know what the context of it is.
`
`19
`
`20
`
`21
`
`22
`
`done.
`
`MR. VIGEN: Okay.
`
`THE COURT: Okay. Show it to him.
`
`You go ahead and read it, and tell us when you're
`
`23
`
`MR. VIGEN: Your Honor, I wanted to -- I want to show
`
`24
`
`the witness Government Exhibit 590, please.
`
`25
`
`THE COURT: 590?
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 13 of 288
` 931
`
`Pierre Cremieux - Cross
`
` 1
`
` 2
`
`MR. VIGEN: Yes, sir.
`
`THE COURT: What's 590?
`
` 3
`
`BY MR. VIGEN:
`
` 4
`
`Q. Sir, this is a textbook that you cited in your expert
`
` 5
`
`opinion; correct?
`
` 6
`
`A. Yes.
`
` 7
`
`Q. Okay. Could we please go to -- you specifically cited
`
` 8
`
`Chapter 5 on differences-in-differences; correct?
`
` 9
`
`A. Yes. I believe that's right.
`
`10
`
`MR. VIGEN: Could we go to the opening paragraph,
`
`11
`
`please.
`
`12
`
`BY MR. VIGEN:
`
`13
`
`Q. Do you see where it says, "This chapter considers a
`
`14
`
`variation on the control theme: Strategies that use data with
`
`15
`
`a time or cohort dimension to control for unobserved but fixed
`
`16
`
`omitted variables. These strategies punt on comparisons and
`
`17
`
`levels while requiring the counterfactual trend behavior of
`
`18
`
`treatment and control groups to be the same."
`
`19
`
`Do you see that?
`
`20
`
`A. I apologize. Can I read it from the beginning, just
`
`21
`
`because I think you started reading and --
`
`22
`
`23
`
`24
`
`sorry.
`
`THE COURT: Absolutely, you can.
`
`THE WITNESS: Okay. Just give me one second. I'm
`
`25
`
`Okay. Thank you.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 14 of 288
` 932
`
`Pierre Cremieux - Cross
`
` 1
`
`BY MR. VIGEN:
`
` 2
`
`Q. Do you agree that the strategies punt on comparisons and
`
` 3
`
`levels while requiring the counterfactual trend behavior of
`
` 4
`
`treatment and control groups to be the same?
`
` 5
`
`A. Yes, I do.
`
` 6
`
`Q. Okay. And the treatment group is the DaVita data; is that
`
` 7
`
`right?
`
` 8
`
`A. That's right.
`
` 9
`
`Q. And the control group is your benchmark data; correct?
`
`10
`
`A. That's correct.
`
`11
`
`Q. Okay. And this is saying that the counterfactual trend
`
`12
`
`behavior of both groups needs to be the same; right?
`
`13
`
`A. That's what this says. Yes.
`
`14
`
`Q. And you agree with that?
`
`15
`
`A. Well, it depends what you applied for. But, yes, I agree
`
`16
`
`generally with the statement. There is nothing wrong with the
`
`17
`
`statement.
`
`18
`
`Q. Okay. I'd like to show what you have done to show that or
`
`19
`
`what you have not done to show that. If we could go to your
`
`20
`
`expert report, Government Exhibit 510. Specifically,
`
`21
`
`Exhibit 3.
`
`22
`
`So the counterfactual trend behavior being the same,
`
`23
`
`you need to show that it's the same before the conspiracy
`
`24
`
`period; you would agree with that?
`
`25
`
`A. That's not quite right.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 15 of 288
` 933
`
`Pierre Cremieux - Cross
`
` 1
`
`Q. Okay. What do you disagree about that?
`
` 2
`
`A. What I disagree about is that what happens within the
`
` 3
`
`pre-conspiracy period is not particularly important. What's
`
` 4
`
`important is that as I compare the levels before and during the
`
` 5
`
`conspiracy period at DaVita, I have an anchor that I can use to
`
` 6
`
`see how that evolution fits within what we see in the more
`
` 7
`
`general economy, in the comparators that DaVita uses for its
`
` 8
`
`compensation.
`
` 9
`
`So that's what I'm doing. I'm literally simply taking
`
`10
`
`a benchmark and -- which is -- which includes DaVita -- the
`
`11
`
`benchmark is the sector of the outpatient healthcare sector --
`
`12
`
`and I'm saying, well, it moved by 12 percent, DaVita moved up
`
`13
`
`by 15 percent, so DaVita moved up faster than the benchmark.
`
`14
`
`That's really beginning and end of the statement I'm making. I
`
`15
`
`made up the numbers; those are not the exact numbers in my
`
`16
`
`report; but, fundamentally, what I'm doing is that.
`
`17
`
`Q. Sir, where did you show that the trend behavior of the
`
`18
`
`DaVita data and the benchmark data was the same prior to the
`
`19
`
`conspiracy?
`
`20
`
`A. I didn't.
`
`21
`
`Q. And that is a necessary requirement -- if we go back to
`
`22
`
`Government Exhibit 590 -- for this to be a valid
`
`23
`
`difference-in-difference approach; correct?
`
`24
`
`A. No, that's not right. You're not reading -- you're
`
`25
`
`understand -- I think you're not understanding that document
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 16 of 288
` 934
`
`Pierre Cremieux - Cross
`
` 1
`
`correctly.
`
` 2
`
`Q. Okay.
`
` 3
`
`A. That's not what it says.
`
` 4
`
`Q. When this document says that, "These strategies punt on
`
` 5
`
`comparisons and levels, while requiring the counterfactual
`
` 6
`
`trend behavior of treatment and control groups to be the same,"
`
` 7
`
`what is that saying you're supposed to show to show a valid
`
` 8
`
`difference-in-difference approach?
`
` 9
`
`A. What that says is that since I'm comparing two periods for
`
`10
`
`DaVita and I'm comparing two periods for the benchmark, the
`
`11
`
`movement between the two absent the allegations should
`
`12
`
`expect -- should be expected to be similar. And since the
`
`13
`
`benchmark is in fact the benchmark that DaVita uses to
`
`14
`
`determine its compensation, that seems like quite a reasonable
`
`15
`
`assumption to me.
`
`16
`
`Q. Okay. But you need to actually show that those trends in
`
`17
`
`the before period are moving in similar fashion; that's what
`
`18
`
`you just said.
`
`19
`
`A. No, that's not what I just said. I'm sorry. You're not
`
`20
`
`understanding. You're making the mistake of conflating what
`
`21
`
`happens before -- in the before period with the difference in
`
`22
`
`the trend before and after. So you cannot just look at one
`
`23
`
`period. Because, again, if you look at my report, you'll see
`
`24
`
`that the -- in the before period -- in fact, you put it up on
`
`25
`
`the chart earlier. If you go to the chart with the percentage
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 17 of 288
` 935
`
`Pierre Cremieux - Cross
`
` 1
`
`change, you'll see that I have something like a 3.6, I think.
`
` 2
`
`Q. 3.9.
`
` 3
`
`A. 3.9 percent change difference. That number in and of
`
` 4
`
`itself is not particularly interesting. What is interesting is
`
` 5
`
`how it compares with the number after.
`
` 6
`
`Q. Right. If you go back to 590, you're saying, what is
`
` 7
`
`interesting is how it compares?
`
` 8
`
`A. You mean 510; right?
`
` 9
`
`Q. No. 590.
`
`10
`
`A. Okay.
`
`11
`
`Q. The counterfactual in your analysis is the before period;
`
`12
`
`would you agree with that?
`
`13
`
`A. No, I don't. That's I think the source of the error. The
`
`14
`
`counterfactual here is not the before period. The
`
`15
`
`counterfactual is the growth rate before and after for the
`
`16
`
`benchmark. I'm not making any statements about the before
`
`17
`
`period being the counterfactual. It's -- methodologically,
`
`18
`
`what I'm doing is different from what I think what you have in
`
`19
`
`your head as my method.
`
`20
`
`Q. The trend behavior between what you're studying, DaVita's
`
`21
`
`data and the benchmark, absent an alleged conspiracy, they need
`
`22
`
`to be moving in a relatively similar trend. That's what the
`
`23
`
`word -- that's what the whole purpose of this trend analogy is;
`
`24
`
`right?
`
`25
`
`A. I agree with that. Yes.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 18 of 288
` 936
`
`Pierre Cremieux - Cross
`
` 1
`
`Q. Okay. Where in your report did you show that the DaVita
`
` 2
`
`data and the benchmark data moved over time in a similar trend?
`
` 3
`
`A. I don't. That's the whole point. Is I'm showing that if
`
` 4
`
`you look at the benchmark -- which, again, is used by DaVita in
`
` 5
`
`its own determination of compensation -- and I compare it with
`
` 6
`
`the trend for DaVita itself, I observe a difference. I'm not
`
` 7
`
`sure what more you'd expect me to do.
`
` 8
`
`Q. When we were talking about all of these variables before
`
` 9
`
`that impact both the benchmark and DaVita --
`
`10
`
`A. Uh-huh.
`
`11
`
`Q. -- the whole assumption of a difference-in-difference
`
`12
`
`approach is that those variables taken together, you don't need
`
`13
`
`to separately analyze, because they are all moving in the same
`
`14
`
`trend absent what you're studying here, the alleged conspiracy;
`
`15
`
`right?
`
`16
`
`A. Over the entire period?
`
`17
`
`Q. Correct.
`
`18
`
`A. Yes. Not in the pre-period, as you've been saying.
`
`19
`
`Q. Well, but it's important also in the pre-period, would you
`
`20
`
`agree, to show a parallel trend?
`
`21
`
`A. Not necessarily. What's important is to show that for the
`
`22
`
`two periods you're looking at, the benchmark is appropriate.
`
`23
`
`And, again, this benchmark, I'm not making it up. I'm taking
`
`24
`
`it from what DaVita uses as its own benchmark when it tries to
`
`25
`
`figure out how much to compensate its employees.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 19 of 288
` 937
`
`Pierre Cremieux - Cross
`
` 1
`
`MR. VIGEN: Could we show you Government Exhibit 589,
`
` 2
`
`please.
`
` 3
`
`BY MR. VIGEN:
`
` 4
`
`Q. This is another document you cited in your report, the
`
` 5
`
`"Proving Antitrust Damages."
`
` 6
`
`A. Uh-huh.
`
` 7
`
`Q. If we can start in the middle paragraph, starting with
`
` 8
`
`"further."
`
` 9
`
`"Further, the benchmark and affected markets should be
`
`10
`
`characterized by sufficiently comparable economic conditions,
`
`11
`
`at least after controlling for observable factors, such that
`
`12
`
`prices in those markets would have been identical had there
`
`13
`
`been no anticompetitive behavior." Do I read that right?
`
`14
`
`A. I'm sorry. You're going -- at least for me, you're going
`
`15
`
`too fast. What are you reading from?
`
`16
`
`THE COURT: I can't cure him of that. I tried.
`
`17
`
`BY MR. VIGEN:
`
`18
`
`Q. Middle way through the first paragraph on the screen there.
`
`19
`
`"Further, the benchmark" --
`
`20
`
`A. Yes.
`
`21
`
`Q. Can you read it --
`
`22
`
`A. Give me one second to read from the beginning, and then
`
`23
`
`I'll be happy to focus on what you'd like me to focus on.
`
`24
`
`25
`
`MR. VIGEN: Could we highlight that "A" please.
`
`THE WITNESS: Okay. I see this.
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 20 of 288
` 938
`
`Pierre Cremieux - Cross
`
` 1
`
`BY MR. VIGEN:
`
` 2
`
`Q. And then if we can go down to the next paragraph, where it
`
` 3
`
`says, "Under the assumption that the differences would have
`
` 4
`
`been stable, controlling for observable factors but for the
`
` 5
`
`alleged anticompetitive conduct, the difference-in-differences
`
` 6
`
`provides an alternative approach to estimating damages."
`
` 7
`
`A. Yeah, I see that.
`
` 8
`
`Q. Okay. And this is generally saying what we have been
`
` 9
`
`discussing before, that the DaVita data and the benchmark data
`
`10
`
`should be moving in relatively stable, similar trends for this
`
`11
`
`to be a valid analysis. Would you agree with that?
`
`12
`
`A. Absent the allegations.
`
`13
`
`Q. Absent the allegations.
`
`14
`
`A. There we have --
`
`15
`
`Q. So let's go back to your Exhibit 3 in Government
`
`16
`
`Exhibit 510. So absent the allegations, we see that in your
`
`17
`
`before period; correct?
`
`18
`
`A. That's right.
`
`19
`
`Q. Okay. But you have not shown across those three years that
`
`20
`
`it is a stable trend, have you?
`
`21
`
`A. Again, I haven't. But you're not understanding, I think,
`
`22
`
`the point of the previous article. That's not what the article
`
`23
`
`is saying. What the article is saying is that over the entire
`
`24
`
`period -- so in this case, 2009 through 2017 or 2019 -- would
`
`25
`
`you expect, absent the allegation, for the trends to be roughly
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 21 of 288
` 939
`
`Pierre Cremieux - Cross
`
` 1
`
`the same for the entire period?
`
` 2
`
`Q. But, sir, you just acknowledged in the before period, when
`
` 3
`
`there is no allegation, that in and of itself that before
`
` 4
`
`period should be stable, as compared to the DaVita data and the
`
` 5
`
`benchmark data.
`
` 6
`
`A. I'm sorry. Should be stable, as compared to the DaVita
`
` 7
`
`data and the benchmark data? I don't know --
`
` 8
`
`Q. The trend should be stable to have a valid
`
` 9
`
`difference-in-difference approach?
`
`10
`
`A. Which trend?
`
`11
`
`Q. The trend in the difference-in-difference between the
`
`12
`
`DaVita compensation data and the benchmark compensation data.
`
`13
`
`A. I'm sorry. I mean, I'm really not understanding. There is
`
`14
`
`no such thing as a trend in the before period in the
`
`15
`
`difference-in-difference. The difference-in-difference
`
`16
`
`analysis is all about using both the before period and the
`
`17
`
`after period -- that's the first difference -- and using the
`
`18
`
`benchmark versus DaVita. That's the second difference in
`
`19
`
`difference-in-difference. So you can't have a
`
`20
`
`difference-in-difference in one of the two periods. It doesn't
`
`21
`
`make sense. You can have a difference, which is 3.9 percent.
`
`22
`
`Q. Sir, before I got to the ultimate question that would show
`
`23
`
`why this is problematic, what you've done, you acknowledge that
`
`24
`
`in the before period, the trend between the DaVita data and the
`
`25
`
`benchmark data should be similar. You acknowledged that right
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 22 of 288
` 940
`
`Pierre Cremieux - Cross
`
` 1
`
`before I got into the follow-up questions; right?
`
` 2
`
`A. I have not looked at that. That's correct.
`
` 3
`
`Q. You haven't looked at it. Okay. We have.
`
` 4
`
`A. Doesn't matter, but that's correct.
`
` 5
`
`Q. We have. I'd like to show you what we did with your data.
`
` 6
`
`THE COURT: Well, you don't show him anything until
`
` 7
`
`you show it to your opposing counsel first.
`
` 8
`
` 9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`Show it to the Court, too, if you wouldn't mind.
`
`MR. VIGEN: Julie, can we have the Elmo?
`
`THE COURT: This is something you created, Mr. Vigen?
`
`MR. VIGEN: From his data.
`
`THE COURT: Right.
`
`MR. VIGEN: Correct.
`
`THE COURT: You don't have an expert; it's you?
`
`MR. VIGEN: Well, this is what I'm representing; and I
`
`16
`
`want to see if he did this. It's a demonstrative.
`
`17
`
`MR. MELSHEIMER: Your Honor, I guess -- I appreciate
`
`18
`
`the lovely burnt orange and the University of Texas homage in
`
`19
`
`this, but I don't understand how he's going to cross-examine --
`
`20
`
`I don't understand how he's going to cross-examine Dr. Cremieux
`
`21
`
`on something that he did without any basis for how it's done or
`
`22
`
`what was done or anything like that.
`
`23
`
`THE COURT: Well, during this so-called Daubert
`
`24
`
`hearing, I'll let him do that. But when it comes to the actual
`
`25
`
`trial in front of the jury, unfortunately for -- maybe
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 23 of 288
` 941
`
`Pierre Cremieux - Cross
`
` 1
`
`fortunately -- for the government, Mr. Vigen can't take the
`
` 2
`
`stand.
`
` 3
`
`MR. MELSHEIMER: Thank you.
`
` 4
`
`BY MR. VIGEN:
`
` 5
`
`Q. Okay. So on your Exhibit 3, you showed an average of the
`
` 6
`
`2009 to 2011 years; is that right?
`
` 7
`
`A. Can you show me Exhibit 3 again?
`
` 8
`
`That's -- that seems -- I mean, again, I don't know
`
` 9
`
`how you generated those numbers; I don't know where they come
`
`10
`
`from; I don't know if they're right; so I'm just going to take
`
`11
`
`your word for it for now. So the 3.9 percentage points is the
`
`12
`
`average over that period. It's the average elevation in
`
`13
`
`job-to-job turnover for the senior-level employees in that
`
`14
`
`period. Yes.
`
`15
`
`MR. VIGEN: If we go back to the demonstrative on the
`
`16
`
`Elmo, please.
`
`17
`
`BY MR. VIGEN:
`
`18
`
`Q. This is the type of analysis you did not show; right? You
`
`19
`
`did not show the trend in the before period; correct?
`
`20
`
`A. That's right. It's not relevant to my analysis.
`
`21
`
`Q. Okay.
`
`22
`
`A. That's why I'm doing the difference-in-difference.
`
`23
`
`Q. The difference-in-difference analysis, as we discussed,
`
`24
`
`requires that your control group, your benchmark, and the
`
`25
`
`DaVita data have a similar trend outside of the effective
`
`
`
`Case 1:21-cr-00229-RBJ Document 291 Filed 06/16/22 USDC Colorado Page 24 of 288
` 942
`
`Pierre Cremieux - Cross
`
` 1
`
`conspiracy wi
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
