Food Lion, LLC, et al v. Dean Foods Company et al

Filing 667

MEMORANDUM OPINION. For the reasons set forth herein, plaintiffs' Objection 630 APPEAL OF MAGISTRATE JUDGE DECISION to District Court, is OVERRULED and the Magistrate Judge's order, 625 , is AFFIRMED AND ADOPTED by the Court and made the ORDER of the Court. Signed by District Judge J Ronnie Greer on 3/20/2012. (FMM )

IN THE UNITED STATES DISTRICT COURT
FOR THE EASTERN DISTRICT OF TENNESSEE
GREENEVILLE DIVISION

IN RE: SOUTHEASTERN MILK
ANTITRUST LITIGATION

THIS DOCUMENT RELATES TO:
Food Lion, LLC., et al. v. Dean Foods,
et al., No. 2:07-CV 188

)
)
)
)
)
)
)

Master File No. 2:08-MD-1000
Judge J. Ronnie Greer
Magistrate Judge Dennis H.
Inman

MEMORANDUM OPINION
This matter is before the Court on the retailer plaintiffs’ objection to the December 8, 2010
order of the Magistrate Judge, [Doc. 1191], granting in part defendants’ motion to exclude the
testimony of plaintiffs’ expert Professor Luke Froeb, [Doc. 1206]. The defendants have responded
to the objection, [Doc. 1229], and plaintiffs have filed a reply, [Doc. 1249] 1. For the reasons which
follow, the objection of the plaintiffs is overruled and the Magistrate Judge’s order, [Doc. 1191], is
AFFIRMED and ADOPTED by this Court and made the order of the Court.
I. Plaintiffs’ Objection to December 8, 2010 Order
Professor Luke Froeb (“Professor Froeb”) is an associate professor of entrepreneurship and
free enterprise at Vanderbilt University. Prior to his work at Vanderbilt, he was director of the
Bureau of Economics at the Federal Trade Commission and an economist with the Antitrust Division
of the Department of Justice. Professor Froeb was retained as an expert witness by plaintiffs to offer
opinions on the relevant geographic and product markets and defendants’ market power. The

1

Defendants’ underlying Daubert motion to exclude the testimony of Professor Froeb and the
supporting memorandum are Docs. 1088 and 1089; plaintiffs’ response is Doc. 1159; and defendants’ reply
is Doc. 1172.

defendants moved to exclude Professor Froeb’s testimony pursuant to Federal Rule of Evidence 702
and Daubert v. Merrill Dow Pharm., 509 U.S. 579 (1993).
Relying on the undersigned’s prior decision concerning Professor Froeb’s opinions in the
context of the motion of all defendants for summary judgment, [Doc. 461], as the law of the case,
the Magistrate Judge granted defendants’ motion as to Professor Froeb’s opinion as to the relevant
geographic markets and market power and denied the motion as to his opinion on the relevant
product market, [Doc. 1191]. Plaintiffs now object to that order, arguing that it “suffers from clear
errors of both procedure and substance.”

Plaintiffs seek, alternatively, to have the matter

resubmitted to the Magistrate Judge for a de novo ruling2 or for the Court to reconsider the merits
of the Daubert motion itself.
Although plaintiffs have not sought reconsideration of the Court’s summary judgment order3,
they do argue that the Court’s finding with respect to Professor Froeb’s testimony is “clear error.”
Specifically, plaintiffs contest the Court’s finding that Professor Froeb’s testimony that he used the
same methodology that he would have instructed his staff to employ at the FTC or DOJ does not
establish “that such methodology is legally acceptable based on existing precedent.” They reason
that since DOJ and FTC jointly developed the “Horizontal Merger Guidelines,” which Professor
Froeb purported to use, he therefore used the very “methodology mandated by the Supreme Court
precedent.” Plaintiffs also criticize the Court for making findings regarding Professor Froeb
“without full briefing on the Daubert issues that defendants’ attack raise.”
Rule 72(a) requires the district judge to “consider timely objections and modify or set aside

2

3

This case is no longer assigned to a Magistrate Judge.
In fact, plaintiffs have advised the Court that they do not intend to seek such reconsideration.

2

any part of the order [of a magistrate judge] that is contrary to law or clearly erroneous.” Fed. R.
Civ. P. 72(a); see also 28 U.S.C. § 636(d)(1)(A). Plaintiffs’ procedural approach to this matter is
puzzling. In reality, it is the prior ruling of this district judge which they challenge, not that of the
magistrate judge. The magistrate judge did not opine as to the correctness of the district judge’s
finding; he simply held, and correctly so, that the ruling 4, right or wrong, was the law of the case
and that he could not reconsider the ruling. As the plaintiffs properly acknowledge, the “Magistrate
Judge did not conduct any independent analysis of Professor Froeb’s opinions.”
Plaintiffs claim that the undersigned’s August 14, 2010 memorandum opinion and order
“contains four clear errors of law, stemming from [a] flawed procedure and exacerbated by
defendants’ misrepresentations.” Those four claimed errors are: (1) the district judge in reality ruled
on a Daubert motion “without proper notice and inadequate record;” (2) Professor Froeb did
consider the necessary “commercial realities” and never admitted he did not do so, contrary to the
Court’s finding; (3) Professor Froeb did not consider only a “single customer” but properly applied
the smallest market principle to all milk purchasers, contrary to the Court’s finding; and (4) that

4

The Magistrate Judge’s order, in sum, merely held that this Court’s prior ruling in connection with
defendants’ motion for summary judgment that Professor Froeb’s methodology did not comply with
applicable Supreme Court standards and was thus inadmissible was the law of the case. In that, the
Magistrate Judge was correct and plaintiffs do not explicitly challenge that holding. They merely argue that
the Magistrate Judge’s conclusion was procedurally flawed because the Court did not rule “on whether
Professor Froeb’s proposed testimony on geographic market complies with the applicable Daubert standard,
with the benefit of full briefing on a complete record.” It is not clear that plaintiffs can properly challenge
the undersigned district judge’s ruling through the mechanism of an objection to the order of the Magistrate
Judge. In essence, plaintiffs complain that the Magistrate Judge did not make independent findings regarding
the merits of this Court’s prior opinions, which they now assert were based on findings that were clearly
erroneous. That is confirmed by plaintiffs’ argument that the Magistrate Judge’s order “was contrary to law
because it accepted as the ‘law of the case’ an opinion that was not based on a complete record and was issued
after the Court rejected retailer plaintiffs’ request to await full Daubert proceedings. [Doc. 1206 at 13].
Nevertheless, retailer plaintiffs “advise the Court that they do not intend to seek reconsideration of the Court’s
dismissal of Counts II-IV of the complaint.”

3

Professor Froeb performed his analysis in the same way he would have done it at the DOJ or FDC
demonstrates that he applied the correct Sixth Circuit and Supreme Court methodology.
With respect to plaintiffs’ objections to the magistrate judge’s order, the Court will discuss
only the first of these claimed errors in this section, as the others go, not to the merits of the
objections, but to the merits of the Daubert motion itself which will be discussed below. Plaintiffs
rely on Jahn v. Equine Services, 233 F.3d 382, 387 (6th Cir. 2000) to support their argument that
the Court made its rulings with regard to Professor Froeb’s testimony “without proper notice and
inadequate record.” In Jahn, a veterinary malpractice case involving the death of a champion
Hackney pony, all defendants moved for summary judgment after completion of discovery. The
motions turned on causation and the district court, sua sponte and without allowing additional
briefing, held that the testimony of Jahn’s expert veterinarian witness was inadmissible under
Daubert. Without the testimony, Jahn could not prove causation and summary judgment was
granted. Although the Sixth Circuit found substantive reasons for reversing the district court, it also
found “that the record was insufficient in order to make a proper Daubert determination.” Id. at 393.
“A district court should not make a Daubert ruling prematurely, but should only do so when the
record is complete enough to measure the proffered testimony against the proper standards of
liability and relevance.” Id.
As pointed out in Jahn, “[t]he district court is not obligated to hold a Daubert hearing,” id.
(quoting Clay v. Ford Motor Co., 215 F.3d 663, 667 (6th Cir. 2000)), when the record is “adequate
to the task.”5 Id. The Jahn opinion does not specifically address the question of what must be in

5

Although plaintiffs now request a hearing, they cannot make any serious argument that the record
is not adequate for the Court to now decide admissibility without a hearing, were it to choose to do so.

4

the record for the record to be adequate; however, the expert reports, pathologist reports, and
depositions of most of those involved, including the expert’s depositions, were deemed insufficient
in that case. Id. at 393. Usually, the record upon which the trial court will make an admissibility
decision without a hearing will consist of the expert reports, affidavits, depositions, briefs of the
parties, and the like.
Plaintiffs argue that as of August 4, 2010, the date of the Court’s order on the motion for
summary judgment, “[t]his case was in about exactly the same procedural posture as Jahn.” They
claim that the issue related to Professor Froeb’s methodology was raised for the first time in a
supplemental pleading filed by defendants and that their ability to respond was limited. They
specifically argue that the Court based its opinion “on an incomplete reading of Professor Froeb’s
deposition testimony, which retailer plaintiffs were able to complete by filing a supplemental
declaration by Professor Froeb (as well as by explaining precisely how Professor Froeb had
conducted his analysis).” Plaintiffs cite several other materials, now in the complete Daubert record,
which were not in the original record prior to August 4, 2010, [see Doc. 1206, pp. 14-15].
Defendants, on the other hand, assert that “the only difference between the record [prior to
August 4, 2010] and now” is Professor Froeb’s “eleventh-hour declaration” which they characterize
as Professor Froeb’s attempt to say “what he really meant” by his deposition responses. They argue,
with relation to this declaration, that “Professor Froeb’s say-so that he applied the [horizontal
merger] guidelines does not make it so.” As an exhibit to their opposition to plaintiffs’ objection
to the magistrate judge’s order, defendants have prepared a “pertinent procedural time line” to
illustrate their argument, [see Doc. 1229, Ex. D].
Although the Court will consider defendants’ Daubert motion anew (see discussion below),

5

it does so only out of an abundance of caution in light of plaintiffs’ claims that the Court “jumped
the gun” and decided the issue related to Professor Froeb prematurely, and, the Court does so
alternatively. On the first issue raised by plaintiffs, the only one which goes to the merits of their
objection to the Magistrate Judge’s order, however, the Court largely agrees with the defendants.
As to the issues related to Professor Froeb’s opinions decided by the Court in its summary judgment
order, the record was complete and the items referred to by plaintiffs add little or nothing to the
Court’s analysis, except for the supplemental declaration of Professor Froeb himself. With respect
to the procedural objections made by plaintiffs that they informed the Court that the matters were
not ripe for resolution and that their response was somehow limited (presumably by the Court), that
is simply not the case.
During a hearing on July 1, 2009, the Court specifically inquired of plaintiffs’ counsel
whether plaintiffs could respond to an anticipated motion for summary judgment. Counsel
responded that he thought they could. Then, on November 5, 2009, plaintiffs “advised the Court that
they did not believe that the record was complete and that they needed Professor Froeb’s report to
oppose defendants’ motion for summary judgment.” 6 Thereafter, the Court did not immediately
rule on the motion for summary judgment but withheld ruling until August, 2010, nine months later,
and then only after allowing the plaintiffs oral argument on the motion for summary judgment,
permitting the filing of numerous replies and supplemental briefs and holding the matter in abeyance
until after the close of all discovery of fact witnesses, the filing of expert reports, expert depositions
and supplementation of the record with Professor Froeb’s full report. In this Court’s view, any

6

Plaintiffs’ response to the motion for summary judgment, [Doc. 538], was filed on November 5,
2009. Attached to the response was the Rule 56(e) affidavit of counsel. Document 538 was subsequently
stricken from the record by the Magistrate Judge and refiled as Document 669 on January 22, 2010.

6

suggestion by plaintiffs that the Court ruled on an incomplete record or denied them adequate time
to fully respond is disingenuous and self-serving. Likewise, their claim that the Court limited them
to a five-page brief is also disingenuous. Plaintiffs never, at any time prior to the instant objection,
asked the Court to permit further briefing or to extend the number of pages allowed.
Finally, plaintiffs’ suggestion that the Court based its opinions on “an incomplete reading
of Professor Froeb’s deposition testimony,” which they were able to cure through the filing of his
supplemental declaration, is also somewhat hypocritical. First, they do not even suggest that
Professor Froeb’s declaration was unavailable prior to its filing on November 15, 2010, or that it
could not have been filed when defendants first raised the issue of the methodology used by
Professor Froeb, citing his deposition testimony and before the Court’s ruling on the summary
judgment motion on August 4, 2010.
Plaintiffs could have easily have filed the affidavit of Professor Froeb long before the Court’s
August 4, 2010 decision on the motion for summary judgment. Professor Froeb was deposed on
April 26, 2010, and plaintiffs served defendants with Professor Froeb’s errata sheet on June 7, 2010.7
And, had plaintiffs done so, the Court could have considered it in connection with its consideration
of the motion, despite defendants’ suggestion that the supplemental motion is simply an effort by
Professor Froeb and plaintiffs to “get a mulligan.”
Although “a party cannot create a genuine issue of fact sufficient to survive summary
judgment simply by contradicting his or her own previous sworn statement (by say, filing a later
affidavit that flatly contradicts that party’s earlier sworn deposition) without explaining the

7

It is not suggested by plaintiffs that Professor Froeb noted any of his November 14 “clarifications”
on the errata sheet.

7

contradiction or attempting to resolve the disparity,” Cleveland v. Policy Mgmt. Sys. Corp., 526 U.S.
795, 806 (1999), that rule does not extend to a non-party witness. See Hanson v. City of Fairview
Park, 349 Fed. App’x 70, 2009 WL 3351751 (C.A. 6 (Ohio)). Even if the rule applied, Professor
Froeb’s supplemental declaration does not flatly contradict his earlier deposition testimony but
merely attempts to clarify or explain it. 8
II. The Daubert Motion
Professor Froeb was retained by plaintiffs for the purpose of rendering an expert opinion on
the relevant product and geographic markets for purposes of this litigation. His opinion as to the
relevant product market is not at issue; the motion challenges his opinions as to the definition of the
relevant geographic market. For purposes of plaintiffs’ conspiracy claim, Professor Froeb defines

8

As noted elsewhere, plaintiffs have not asked the Court to reconsider its prior summary judgment
ruling and the dismissal of certain counts of the complaint because of plaintiffs’ inability to establish the
relevant market. If plaintiffs had asked for reconsideration, the Court would be well justified in refusing to
consider Professor Froeb’s supplemental declaration. See Sommer v. Davis, 317 F.3d 686 (6th Cir. 2003)
(holding that a district court’s refusal to consider evidence produced for the first time on a motion to
reconsider will not be reversed absent an abuse of discretion).
Secondly, and more fundamentally, however, plaintiffs attempt to explain Professor Froeb’s
deposition testimony through a supplemental declaration is misplaced. Professor Froeb’s deposition
testimony is what it is. While there might arise a situation where it would be necessary and appropriate for
a witness to explain or clarify deposition testimony, this is not such a case. This is not a situation where
Professor Froeb misspoke or gave an incomplete response. This is a situation where he attempts, through
his supplemental declaration, to “clarify,” unambiguous testimony about the methodology he used, e.g., an
effort to explain “what he really meant” by his answer, as defendants argue. The supplemental declaration
adds nothing of substance to the Court’s inquiry. Indeed, the practice of allowing witnesses to endlessly
explain or clarify their prior testimony, depending on how it was interpreted, would undermine the utility of
the summary judgment procedure and needlessly prolong litigation.
As set forth above, the plaintiffs’ procedural objection to the Magistrate Judge’s order based on Jahn
is not well taken and thus the Magistrate Judge’s holding that the Court’s prior ruling with respect to
Professor Froeb’s opinion is the law of the case must be affirmed. For the reasons already mentioned, and
because exclusion of Professor Froeb’s opinion may result in a grant of summary judgment on Count V and
on plaintiffs’ rule of reason claim in Count I, the Court will, alternatively, although it is not necessary to do
so, revisit the question of the admissibility of Professor Froeb’s testimony.

8

the relevant geographic market as Federal Milk Market Orders 5 and 7. Federal Order 5 includes
North Carolina, South Carolina and parts of Georgia, Indiana, Kentucky, Tennessee, Virginia and
West Virginia. Federal Order 7 includes Alabama, Arkansas, Louisiana, Mississippi, the rest of
Georgia and Tennessee, and parts of Florida, Kentucky and Missouri. For purposes of plaintiffs’
monopolization claims, Professor Froeb defines the relevant geographic market as North Carolina,
South Carolina, Virginia, Georgia and Eastern Tennessee.
In general, defendants argue that Professor Froeb’s market opinions are based on the use of
a theoretical model inconsistent with the applicable legal standards and based on assumptions, not
real world facts. More specifically, defendants claim that Professor Froeb’s opinions are based on
a theoretical model constructed for the purpose of his analysis in this case, a model he characterizes
as an application of the “hypothetical monopolist” or “SNNIP” test utilized by the DOJ/FTC as set
forth in their Horizontal Merger Guidelines.
Plaintiffs respond that Professor Froeb applied the SNNIP test, a valid and accepted
methodology, and that the test, “by its very definition,” complies with Supreme Court dictates on
definition of the relevant market, as well as relevant Sixth Circuit analysis. They further argue that
Professor Froeb captured “appropriate ‘market realities,’” such as transportation costs, locations of
milk plants within Orders 5 and 7 and actual consumer locations. Plaintiffs accuse defendants of
a “purposeful misconstruction” of Professor Froeb’s deposition testimony on these subjects.
1.

The Daubert Standard

The starting point for determining the admissibility of expert testimony in federal courts is
Federal Rule of Evidence 702. Rule 702 provides:
If scientific, technical, or other specialized knowledge will
assist the trier of fact to understand the evidence or to determine a
9

fact in issue, a witness qualified as an expert by knowledge, skill,
experience, training, or education, may testify thereto in the form of
an opinion or otherwise, if (1) the testimony is based upon sufficient
facts or data, (2) the testimony is the product of reliable principles
and methods, and (3) the witness has applied the principles and
methods reliably to the facts of the case.
Fed. R. Evid. 702
This Rule reflects the standards enunciated in the Supreme Court’s decisions in Daubert v.
Merrill Dow Pharm., Inc., 509 U.S. 579 (1993) and Kumho Tire Co. v. Carmichael, 526 U.S. 137
(1999). Fed. R. Evid. 702 advisory cmte. notes, 2000 Amend. (“In Daubert the court charged trial
judges with the responsibility of acting as gatekeepers to exclude unreliable expert testimony, and
the court in Kumho clarified that this gatekeeper function applies to all expert testimony, not just
testing based in science.”). Under the Rule, a proposed expert’s opinion is admissible, at the
discretion of the trial court, if the opinion satisfies three requirements. First, the witness must be
qualified by “knowledge, skill, experience, training, or education.” Second, the testimony must be
relevant and “assist the trier of fact to understand the evidence or to determine a fact in issue.”
Third, the testimony must be reliable. Fed. R. Evid. 702.
Rule 702 requires an expert witness to base his opinions upon “sufficient facts and data,” rely
on “reliable principles and methods,” and apply “the principles and methods reliably to the facts of
the case.” Id.; see also Fed. R. Evid. 702 advisory cmte. notes, 2000 Amend. (“[A]ny step that
renders the analysis unreliable . . . renders the expert’s testimony inadmissible.”) (internal quotation
marks omitted). But, “rejection of expert testimony is the exception, rather than the rule.” Id.
Generally, if an expert’s opinion amounts to “mere guess or speculation,” it should be
excluded. United States v. L.E. Cooke Co., 991 F.2d 336, 342 (6th Cir. 1993). Even though
“shaky,” however, opinions based on facts in the record are admissible and challenges merely go
10

to the accuracy of the conclusions rather than the reliability of the testimony. Jahn, 233 F.3d at 39093. “An expert’s opinion, where based on assumed facts, must find some support for those
assumptions in the record. However, mere ‘weaknesses in the factual basis of an expert witness’
opinion . . . bear on the weight of the evidence rather than on its admissibility.’” McLean v. 988011
Ontario, Ltd., 224 F.3d 797, 801 (6th Cir. 2000) (quoting Cooke, 991 F.2d 342) (internal citations
omitted).
The gatekeeper inquiry under Rule 702 and Daubert is a flexible determination, allowing the
district court to exercise discretion. Kumho Tire, 526 U.S. at 157-58. Daubert provides a nonexclusive checklist for trial courts in evaluating the reliability of expert testimony that must be
tailored to the facts of a particular case. Id. at 150 (citing Daubert, 309 U.S. at 593). Those factors
include: testing, peer review, publication, error rates, the existence and maintenance of standards
controlling the technique’s operation, and general acceptance in the relevant scientific community.
Daubert, 509 U.S. 593-94. The Daubert factors “are not dispositive in every case” and should be
applied only “where they are reasonable measures of the reliability of expert testimony.” Gross v.
Comm’r., 272 F.3d 333, 339 (6th Cir. 2001).
2.

Analysis and Discussion

The relevant antitrust geographic market has been defined by the United States Supreme
Court as “the region in which the seller operates, and to which the purchaser can practicably turn
for supplies.” Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 327 (1961). The relevant
geographic market must both “correspond to the commercial realities of the industry and be
economically significant.” Brown Shoe Co. v. United States, 370 U.S. 294, 336-37 (1962); Tampa
Elec., 365 U.S. at 327.

11

To facilitate the analysis in the context of horizontal acquisitions and mergers, the
Department of Justice and Federal Trade Commission established the Horizontal Merger Guidelines
(“Merger Guidelines”) which describe the “analytical framework and specific standards normally
used” by them in analyzing mergers. Merger Guidelines, § 0. The Merger Guidelines define the
geographic market as “a region such that a hypothetical monopolist that was the only present or
future producer of the relevant product at locations in that region would profitably impose at least
a ‘small but significant and non-transitory’ [“SNNIP”] increase in price, holding constant the terms
of sale for all products produced elsewhere.” Merger Guidelines, § 1.21. If buyers would respond
to the SNNIP by shifting to products produced outside the proposed geographic market, and this
shift were sufficient to render the SNNIP unprofitable, then the proposed geographic market would
be too narrow. Id.
The Sixth Circuit has acknowledged that the use of the SNNIP test “has been recognized in
antitrust case law.” Kentucky Speedway, LLC v. NASCAR, Inc., 588 F.3d 908 (6th Cir. 2009) (citing
FTC v. Whole Foods Mkt., Inc., 548 F.3d 1028, 1038 (D.C. Cir. 2008) (recognizing the SNNIP test
as a valid diagnostic tool) (emphasis added)). So far as this Court can tell, however, neither the
United States Supreme Court nor the Sixth Circuit have specifically equated the SNIPP test, or any
variation of it, with the standard enunciated in Tampa Electric. More specifically, the Sixth Circuit
has not, as claimed by plaintiffs, held that “[t]he SNIPP test applied by Professor Froeb is the law
of this circuit.” [Doc. 1159 at 2]. The parties in this case do, however, acknowledge that the SNIPP
test is an accepted method of analyzing the geographic market in this antitrust case. The real
question for the Court then is whether the methodology employed by Professor Froeb, by whatever

12

name it is labeled, complies with the standard enunciated by the Supreme Court.9
Perhaps the best indication of the methodology used by Professor Froeb in formulating his
opinions as to the relevant geographic market is his own description of the methodology. He has
provided four declarations in the case and has been deposed extensively. His declarations are dated
November 5, 2009 (offered in support of plaintiffs’ response to defendants’ motion for summary
judgment), March 5, 2010 (his expert report), June 28, 2010 (his rebuttal report) and November 14,
2010 (clarifying his deposition testimony). Professor Froeb was deposed on April 26, 2010.
In the first declaration, Professor Froeb declares that he has “begun the process of
constructing an economic model of competition among bottling plants and the effects of that
competition on processed milk customers.” [¶ 4]. He quotes the definition of an antitrust market
from the Merger Guidelines and opines that the “hypothetical monopolist test is an appropriate
paradigm to delineate a relevant market.” [¶ 8]. Professor Froeb claims that defendants’ “use of
shipment data to critique Plaintiffs’ proposed market is a well known fallacy for market delineation.”
[¶ 10]. He further opines that “[m]arket delineation is a fact intensive inquiry.” [¶ 14].
In the second declaration, Professor Froeb sets out his conclusions as to the relevant markets
and states that he “appl[ies] the methodology of the U.S. Department of Justice and Federal Trade
Commission Horizontal Merger Guidelines to identify a group of products, and the area in which
they are produced or consumed, over which market power could be exercised.” [¶ 8]; see also ¶ 30
(. . . “I use the hypothetical monopolist paradigm to delineate a relevant market.”). He states his
opinions and conclusions about the relevant product market and the relevant geographic markets for
9

Professor Froeb acknowledged in his deposition that he was “not aware of any specific court
decisions” that have allowed the use of the model used by him in this case in a non-merger antitrust case.
Likewise, plaintiffs have cited no case where Professor Froeb’s theoretical model has been specifically
approved for use in determining the relevant market.

13

the conspiracy and monopolization claims and Dean’s market power as follows:
10.
My conclusions regarding market definition are the
following: First, I conclude that fresh white fluid milk is a relevant
product market. I base this conclusion on the unique properties of
milk and the inelastic demand for the product. Inelastic demand
means that following a price increase, few consumers would
substitute to other products outside the product market. The less
elastic is demand, the more a hypothetical monopolist, who
eliminates competition among the products in the relevant market,
would increase price.
11.
Second, following the principle that market
delineation should inform the theory of the case, I conclude that there
are two relevant geographic markets related to two of Plaintiffs’
claims. For the claim that defendant Dean Food Company (“Dean”)
monopolized the market for fresh fluid white milk, I conclude that the
relevant geographic market consists of North Carolina, South
Carolina, Virginia, Georgia, and the eastern part of Tennessee. For
the claim that all defendants conspired to raise price to retail
customers, I conclude that the combined region of Federal Orders
Nos. 5 and 7 is a relevant geographic market.
12.
I conclude that the scope of the relevant geographic
market differs for each of the two claims because the goal of market
definition is to inform the theory of the case. A claim of
monopolization should consider the region where milk is sold by the
Defendants and bought by the Plaintiffs. The region should be large
enough to allow a hypothetical monopolist over all plants in the
region profitably to increase price by a significant amount. Based on
my analysis, I conclude that a hypothetical monopolist over all the
plants in North Carolina, South Carolina, Virginia, Georgia, and the
eastern part of Tennessee would find it profitable to raise price by
about 10 percent relative to an industry of independently owned
plants. A claim of coordinated action, on the other hand, should
begin with a candidate market that includes representative plants
owned by the firms in question. Based on my analysis, I conclude
that a hypothetical monopolist controlling all the plants in Orders 5
and 7 would find it profitable to raise price by about 15 percent
relative to an industry of independently owned plants. This implies
that the milk processing plants in Orders 5 and 7 constitute a relevant
geographic market for the purposes of evaluating a conspiracy claim.
13.

For the monopolization claim, I also asked whether
14

Dean has the ability to exercise market power over the consumers in
the relevant market. I conclude that Dean’s control over a significant
number of plants would allow it to raise price by about nine percent
in North Carolina, South Carolina, Virginia, Georgia and the eastern
part of Tennessee. Therefore, Dean has the ability to monopolize the
relevant market.
[March 5, 2010 Declaration, ¶¶¶¶ 10, 11, 12, 13]. He further states that his conclusions “are based
on an economic model that relies on estimates of the demand elasticity for milk and the economic
cost of transporting milk from producers to consumers.” These estimates are a range which,
according to Professor Froeb, “reflects a reasonable range of assumptions regarding how customers
respond to price changes and a reasonable range of measures of the cost of transporting bottled
milk.” [¶ 15].
Professor Froeb further states that he constructed an economic model of competition between
milk producers and “ask[s] whether a hypothetical monopolist of the plants in a candidate market
would find it profitable to raise price by a small but significant amount (e.g. five percent). If not,
then the geographic area is not a relevant antitrust market, and must be increased by adding the nextclosest plants and then re-tested.” [¶ 33]. Professor Froeb’s report gives considerable detail about
the methods by which he reached his conclusions and describes how he performed his hypothetical
monopolist test. “To perform the hypothetical monopolist test, it is first necessary to characterize
competition among milk processors in this industry, and then to determine how a hypothetical
monopolist would eliminate competition among commonly-owned milk processors.” [¶ 52].
Professor Froeb developed an economic model which he used “to predict how much a hypothetical
monopolist would raise price.” [¶ 63]. Professor Froeb’s conclusions are all couched in terms of the
hypothetical monopolist. He “conclude[s] that a hypothetical monopolist of all milk-bottling plants
in Federal Orders Nos. 5 and 7 would find it profitable to impose a price increase of at least five
15

percent.” [¶ 78]. As a result, he concludes that Orders 5 and 7 “is an area over which market power
could be exercised” for the purpose of Plaintiffs’ concerted action claim. [¶ 85].
With respect to plaintiffs’ monopolization claim, Professor Froeb states that he follows the
“methodology of the Horizontal Merger Guidelines,” begins with a relatively small geographic
candidate market, and concludes that a hypothetical monopolist over the entire state of Georgia, or
North Carolina or Virginia “controls too small an area to render a significant price increase
profitable.” [¶ 86]. Based on an analysis “motivated by including regions where Dean and Food
Lion engage in the sale and purchase of milk,” Professor Froeb concludes “a relevant antitrust
market consists of North Carolina, South Carolina, Virginia, Georgia and the eastern part of
Tennessee.” [¶¶ 87, 88]. He further concludes “that a hypothetical monopolist over the relevant
market could profitably raise price approximately 10 percent.” [¶ 89].
So far, so good for plaintiffs at least as to how Professor Froeb characterizes his own work
in the declarations. That, however, is where the puzzle becomes more challenging. Defendants
deposed Professor Froeb on April 26, 2010. His deposition testimony is problematic. Professor
Froeb was asked twice during the deposition whether his market analysis considered “the area in
which the seller operates and to which the purchaser can practicably turn for supplies,” i.e. the
Tampa Electric standard. The relevant testimony is as follows:
Q.

Would you agree that a relevant geographic market is
the area in which the seller operates and to which the
purchaser can practically turn for supplies?
MR. STENERSON: Object to the form.
THE WITNESS:
That’s a different– that’s not
how I would put it. I clearly lay out the principles of
market delineation in my– in my report. And based on
those principles, that led me to the relevant geographic
16

market that– and that I’m looking at. I’m looking at
an area over which a hypothetical monopolist has the
potential to eliminate competition and to raise price.
That’s how I delineated the relevant market.
MS. FEENEY:
Q.

Do you disagree with that statement I just made, that
a relevant geographic market is the area in which the
seller operates and to which the purchaser can
practically turn for supplies?
MR. STENERSON: Object to the form.
THE WITNESS:
For the purposes of informing
the theory of the case, I don’t think that characterizes
well what I’ve done here, or my conclusions here.
And I’ve taken a different approach.

[Doc. ____, Ex. 2, p. 48 ll. 16-24–p. 49, ll 1-16].
Plaintiffs and defendants offer widely different interpretations of this testimony. One thing
is clear, however; it is the Tampa Electric standard enunciated by the Supreme Court which controls
this Court’s evaluation of the admissibility of Professor Froeb’s testimony. Unless and until the
Tampa Electric standard is repudiated or modified by the Supreme Court, this Court is bound to
apply it, and, although the parties agree that the SNNIP test mirrors the Tampa Electric standard,10
neither the Supreme Court nor the Sixth Circuit has explicitly said so,11 plaintiffs’ reading of the
Kentucky Speedway opinion notwithstanding. The Court accepts, however, for the purposes of

10

Plaintiffs refer to the SNNIP test and Tampa Electric as “one and the same.” Defendants agreed
at oral argument on their Daubert motion that the SNNIP test is the proper test, but argue that it was not
properly applied by Professor Froeb.
11

Some courts have in fact recognized that the Merger Guidelines are not binding on the courts.
See Chicago Bridge & Iron Co. v. F.T.C., 534 F.3d 410, 432 FN11 (2008) (“Merger Guidelines are often used
as persuasive authority when deciding if a particular acquisition violates anti-trust laws.”) (citing United
States v. Kinder, 64 F.3d 757, 771 (2d Cir. 1995)). See also Chicago Bridge, 534 F.3d at 434, FN13 and
cases collected therein.

17

deciding this Daubert motion, that proper application of the SSNIP test meets the necessary
standard.
The relevant question, therefore, is whether Professor Froeb defines the relevant geographic
market as “the area in which the seller operates, and to which the purchaser can practicably turn for
supplies.” 12 If the analysis done by Professor Froeb applies that standard, whether referred to as
“SNIPP,” “hypothetical monopolist,” or some variation of that, then it is admissible. Otherwise, it
is not. Defendants accuse Professor Froeb of giving lip-service to the Merger Guidelines, which,
according to plaintiffs, are synonymous with the Supreme Court standard, but in reality using a
theoretical model (i.e. his own version of the SNIPP test) constructed solely for the purpose of this
litigation.
On the basis of Professor Froeb’s sworn deposition testimony, this Court has to agree with
defendants. When asked if he applied the appropriate standard his answer was quite clear. Asked
about whether he delineated the relevant geographic market using the standard enunciated by the
Supreme Court, using language which tracks the Court’s language precisely, Professor Froeb
responded that that description was “not how [he] would put it” but rather that he considered the
“area over which a hypothetical monopolist has the potential to eliminate competition and to raise
prices.” When asked the second time, he was even clearer. “I don’t think [the Tampa Electric
standard] . . . characterizes well what I’ve done here, or my conclusions here . . . . I’ve taken a
different approach.”

Taking his words at face value then, Professor Froeb took a “different

approach” to market definition than the one dictated by the Supreme Court. That “different
12

Professor Froeb futher muddies the water by his statement in his supplemental declaration that
he performed his “economic analysis to address the question of where customers ‘can practically’ turn for
supplies,” omitting any mention of the rest of the Tampa Electric standard. This further illustrates the fact
that clarifying declarations may themselves raise questions which in turn require further clarification.

18

approach” renders his methodology–his theoretical model–inadmissible.
Plaintiffs make several arguments in an effort to salvage Professor Froeb’s testimony. First,
they point to Professor Froeb’s deposition testimony that he would have instructed his employees
at DOJ/FTC to apply the same methodology he applied as establishing that he in fact applied the
Merger Guidelines. This Court previously rejected that testimony as establishing that he did so, a
position the Court does not retreat from. Proof that Professor Froeb would have instructed his
employees to use the same methodology he has now used proves only that, not that the methodology
complies with binding Supreme Court precedent.
Secondly, plaintiffs seem to suggest that Professor Froeb was somehow tricked by
defendants’ counsel at the deposition because the premise of counsel’s questions to Professor Froeb
was not identified as having come from the Tampa Electric case, pointing out that he is not a lawyer.
The Court also rejects this argument. It is ludicrous for plaintiffs to suggest that an expert witness
on the relevant antitrust market would not be familiar with Supreme Court pronouncements on the
subject, especially given his background at DOJ/FTC. The assertions especially lack credibility
when one looks at the reports of Professor Froeb, throughout which he cites or relies upon applicable
legal principles in the antitrust field. For instance, in footnote 3 of Professor Froeb’s report, he
parrots plaintiffs’ argument that they need not prove a relevant geographic market for their
conspiracy claim because “to the extent Plaintiffs claim that Defendants’ conduct was illegal per se,
no market definition is necessary.” He also cites the “Cellophane fallacy,” citing United States v.
E.I. DuPont deNemours & Co., 351 U.S. 377 (1956), affirming 118 F.Supp. 41 (D. Del. 1953).
Most significantly, however, plaintiffs rely in large part on Professor Froeb’s November 14,
2010 declaration which purports to “clarify” his deposition testimony which he claims defendants

19

have “misconstrued.”13 He declares that the “different approach” he referred to “was precisely the
hypothetical monopolist test,” and points out that he notes in his report that he “consider[ed] the
region where milk is sold by the defendants and bought by the plaintiffs.”14 Even if the filing of the
clarifying declaration is appropriate, plaintiffs cannot create an issue where none otherwise existed
by providing a witness’s “clarifying,” after the fact statements. As noted above, the clarifying
declaration sheds little light on Professor Froeb’s testimony that he used a “different approach” than
the one required by Tampa Electric.
To paraphrase Professor Froeb’s deposition testimony, clarified by his declaration: “I would
not say that a relevant geographic market is the area in which the seller operates and to which the
purchaser can practically turn for supplies. That’s not how I would put it. I don’t think the Tampa
Electric standard characterizes well what I’ve done here, or my conclusions here. I’ve taken a
different approach. The different approach I have taken is that I implement the Supreme Court’s
guidance through the use of the hypothetical monopolist test”. Professor Froeb never explained how
the “same approach” can be a “different approach.” The Court will not speculate about the
methodology used by Professor Froeb nor will it attempt to ferret out what he really did (i.e. the
methodology used) without reference to what he said he did. The plaintiffs, as the parties seeking
to have Professor Froeb’s testimony admitted, bear the burden of showing “that the expert’s findings
are based on sound science” and this requires an “objective independent validation of the expert’s

13

Plaintiffs accuse defendants of attempting to “spin” Professor Froeb’s deposition responses and
suggest “his answer does not and cannot change what he actually did do.” In other words, it appears that
plaintiffs invite the Court to ignore what Professor Froeb said he did, and instead find that he used an
appropriate method.
14

So far as the Court can determine, this reference, in paragraph 12 of Professor Froeb’s report, is
the only one that even approximates the standard enunciated in Tampa Electric.

20

methodology.” Daubert v. Merrill Dow Pharmaceuticals, Inc. (on remand), 43 F.3d 1311, 1316 (9th
Cir.), cert. denied, 516 U.S. 869 (1995) (quoted by Smelcer v. Norfolk Southern Ry. Co., 105 F.3d
299, 303 (6th Cir. 1997) abrogated on other grounds by General Elec. Co. v. Joiner, 522 U.S. 136
(1997). [“T]he expert’s bold assurance of validity is not enough.” Id. Furthermore, the court’s
focus is on the expert’s methodology, not the correctness of his conclusions. Id. at 1318 (cited by
Smelcer, 185 F.3d at 303). Plaintiffs have not validated Professor Froeb’s methodology and his
opinions are inadmissible for that reason alone.
Plaintiffs also take issue with the Court’s statements in its August 4, 2010 opinion related
to the requirement, set forth in Tampa Electric, that determination of the relevant antitrust
geographic market requires an assessment of the “commercial realities” facing buyers and sellers
in the market place, Tampa Electric, 365 U.S. at 327, a concept explicitly recognized by the Merger
Guidelines. See Merger Guidelines §§ 1.0, 1.2. While plaintiffs do not contest the Court’s
statement of the applicable standard, they do argue that the following finding was clear error:
Professor Froeb admits that he did not assess the ‘commercial
realities,’ Id., but rather relied solely on his theoretical model. Such
an approach may be academically acceptable; it does not, however,
comply with the Supreme Court’s dictates with respect to
construction of the relevant geographic market.
[Doc. 863 at 29].
Plaintiffs are puzzled by the Court’s use of “Id.” in the first sentence quoted above and
suggest that the defendants “[b]y putting ‘commercial realities’ in quotes [in their supplemental
reply] may have misled the Court into believing that the words were spoken by Professor Froeb.”
[Doc. 1206 at 20]. While the “Id.” reference does appear to have been misplaced, the Court was not
misled by defendants’ conduct or by any “false representation” by defendants that the commercial

21

reality not considered by Professor Froeb had to do with milk supply movement from outside the
candidate market. In support of their position, plaintiffs point to Professor Froeb’s report which
states that Professor Froeb “employ[s] the Cournot model of competition” and “takes into account
the geographic location of existing milk bottling plants in Federal Order No. 5 of milk bottling plants
that compete for customers in Federal Orders Nos. 5 and 7.” [Froeb declaration, March 5, 2010, ¶
64]. He further states that his analytical model takes as input “[e]very non-captive plant in Orders
5 and 7 operating in December, 2008" and “every regulated, non-captive plant within 300 miles of
Orders 5 and 7.”

15

[Id. at ¶ 75]. In short, plaintiffs argue that “Professor Froeb did not ignore

‘commercial realities,’ much less admit to doing so, and the Court’s finding on this point is in error.”
[Doc. 1206 at 21].
Whether the phrasing of the Court’s statement was technically accurate or not, the fact
remains that Professor Froeb does acknowledge that he did not generally consider certain facts
relating to actual behavior within the market, what the Court would characterize as “commercial
realities.” He did not consider or find relevant, for instance, data or materials concerning Food Lion,
Fidel Breto, or any other retailer’s purchasing behavior or the sales or pricing data of Dean or other
milk processors; he did not consider relevant information about where retailers actually turn for
supplies currently; nor did he consider relevant information about how purchasers have actually
reacted to actual price increases. Although Professor Froeb states in his declaration, generically
speaking, that he reviewed “documents produced by the parties in discovery,” there is a paucity of
support in this record for a conclusion that he based his opinions in the case on the evidence in the

15

As discussed below, however, that appears to be contradicted by Professor Froeb’s statement that
his “analysis [for analyzing the monopolization claim] is motivated by including regions where Dean and
Food Lion engage in the sale and purchase of milk.” [Froeb declaration, March 5, 2010, ¶ 87].

22

record but rather that he constructed a theoretical economic model which relied on “estimates” and
“assumptions.” [Froeb declaration, March 5, 2010, ¶ 15].
Two other things illustrate that Professor Froeb largely ignored the commercial realities.
First, plaintiffs identify in their objection two pieces of economic literature which they claim were
necessary for the Court to have a complete Daubert record. Both illustrate how Professor Froeb’s
approach is divorced from actual economic realities. The first is an article written by Gregory J.
Werden, an economist in the Antitrust Division of the Department of Justice.16 [Doc. 1159, Ex. 4].
Plaintiffs characterize the import of the article thusly: “The test must use as ‘inputs’ facts that would
exist in this hypothetical world which are not necessarily observable in the real world.” [Id. at 6].
While acknowledging that the law requires that these fact inputs take into account commercial
realities, plaintiffs claim that “the economist’s job is to figure out how to appropriately take the
necessary realities into account in an exercise that is, by definition, hypothetical.” [Id. at 7]. The
second article, not directly related to the antitrust issues in this case, [Doc. 1159, Ex. 7], notes the
“widely used Cournot model of oligopoly,” a model used by Professor Froeb and one used by the
Federal Energy Regulatory Commission “as a screening device for evaluating a pipeline’s market
power.” Interestingly, the article deals with the “assumptions underlying the Cournot model” and
“standard assumptions used in textbook exposition of the Cournot game.”
Secondly, however, is the testimony of Professor Froeb himself. When questioned about his
model, Professor Froeb gave the following testimony:
BY MS. FEENEY:
Q.

16

The hypothetical monopolist is a construct used for

This article expressly disclaims that its views reflect those of the Department of Justice.

23

purposes of this hypothetical experiment, correct?
A.
It’s an analytical tool used by economists to determine
whether or not there is the potential for the exercise of market power
in the relevant market.
Q.

And it’s generally counterfactual, right,

A.
Counterfactual, you mean by – what do you mean by
counterfactual?
Q.

It’s a hypothetical. It’s not based on real world facts?
MR. STENERSON:

Object to the form. Mischaracterizes his model in his report
and his testimony.
THE WITNESS:
Yeah. So I’m not trying to say
anything about how are people currently behaving in the market. I’m
trying to say: I’m just trying to identify an area in which there is
potential for the exercise of market power.
....
BY MS. FEENEY:
Q.
Would you agree that if a model doesn’t fit the
significant facts, it can result in misleading predictions?
MR. STENERSON: Object to the form.
THE WITNESS:
So models are tools. They are
designed to capture the significant features of competition and predict
how certain actions change that competition. They can–they can give
misleading answers, yes.
BY MS. FEENEY:
Q.
Would you agree that assumptions or predictions used
in a model can be refuted by evidence?
MR. STENERSON: Object to the form.

24

THE WITNESS: If a model doesn’t capture the
significant features of competition, then its predictions may be
misleading.
BY MS. FEENEY:
Q.
So the answer to my question is, yes, you would agree
that assumptions or predictions of a model can be refuted by
evidence?
MR. STENERSON: Object to the form.
THE WITNESS:
That’s not what I said. I said,
look, a model is a tool, and it’s appropriate for some uses, and
inappropriate for others. And what I would – what I would say is that
if a model can capture the significant features of competition in an
industry, then it can be used – or if a model misses significant
features of an industry, and significant features of competition, then
it may give misleading predictions.
[Froeb deposition, April 26, 2010, p. 80, l. 5-81, 1. 1; p. 84, l. 20-86, l. 2]. So, whether you
characterize Professor Froeb’s statements as an “admission” that he did not consider actual
commercial realities or not, Professor Froeb did clearly say he was not “say[ing] anything about how
are people currently behaving in the market.”
Next, plaintiffs take issue with the Court’s finding that Professor Froeb had constructed his
model with reference to a single customer, Food Lion, in violation of applicable legal requirements.
[Doc. 863 at 29]. Plaintiffs incorrectly identify the origin of that conclusion as defendants’
supplemental reply. Rather, the conclusion comes directly from Professor Froeb’s declaration,
where he states: “My analysis is motivated by including regions where Dean [sells] and Food Lion
[purchases] milk.” [Froeb declaration, March 5, 2010, ¶ 87]. Plaintiffs accuse the Court of
misconstruing the language in Professor Froeb’s declaration, but they do not quarrel with the legal
principle underlying the Court’s finding. As the Court stated in its prior memorandum, “A

25

geographic market cannot ordinarily be defined by reference to a single customer.” [Doc. 863 at 29]
(citing Apani Southwest Inc. v. Coca-Cola Enterprises, Inc., 300 F.3d 620, 632-33 (5th Cir. 2002)).
See also Bailey v. Allgas, Inc., 284 F.3d 1237, 1249 (11th Cir. 2002) (“the law is clear, however, that
a geographic market cannot be drawn simply to coincide with the market area of a specific
company”).
Likewise, plaintiffs don’t really dispute that Professor Froeb did not consider the location
of any other retailers, including the other named plaintiff, Fidel Breto. What they do argue is that
Professor Froeb analyzed all purchasers of milk in the market, i.e., people who buy milk, by
considering “population census data to determine where people live as a proxy for demand for milk”
and “assum[ing] that population density would mimic demand for milk, and that Food Lion and
other retailers will be set up to efficiently serve the milk drinking population.” While the argument
has some logical appeal on the surface, it confuses the retail purchasers of milk with the retailers
who purchase milk from processors. It is Food Lion’s purchase of milk from processors which is
relevant and where it purchases milk may be completely unrelated to population census data. That
is further support for defendants’ argument that Professor Froeb’s model is disassociated from the
commercial realities, as defendant suggest.
Finally, plaintiffs argue with that the Court’s previous finding that Professor Froeb’s
assertion that he would have instructed staff at the Department of Justice and Federal Trade
Commission to employ the same methodology he has employed in this case established “only that,”
i.e. that he would have instructed his staff to do so, and does not establish that the methodology
complied with Supreme Court dictates. Plaintiffs claim that Professor Froeb’s assertion, indeed,
does prove “that he applied the standard mandated by the Sixth Circuit” rather than a special test

26

employed for this litigation. The Court continues to disagree, largely for the reasons discussed
above. Accepting that the standards of Tampa Electric and the Merger Guidelines are equivalent,
Professor Froeb must still properly apply the Merger Guidelines and the fact that he might have
instructed staff hundreds of times to apply the methodology he applied in this case simply does not
prove the proposition plaintiffs suggest. In fact, Professor Froeb’s own testimony suggests that he
did not apply the Merger Guidelines, in total, to his work in this case. He testified:
THE WITNESS:
. . . I’ve done this for hundreds
of industries and hundreds of merger cases at the FTC and DOJ.
And that the–the basic methodology that I’m using here,
except I’m applying it to industry of independently owned firms
relative to a hypothetical monopolist.
[Froeb deposition, April 26, 2010, p. 77, ll. 18-24] (emphasis added). In other words, it appears that
Professor Froeb, by his own admission, modified what he had done “for hundreds of industries and
hundreds of merger cases” in applying his model in this case.
As the above illustrates, even if the Court were to grant plaintiffs a full Daubert hearing and
oral argument in the case, there are fundamental reasons why Professor Froeb’s testimony does not
meet the applicable standards and would be inadmissible. Thus, the Magistrate Judge’s order would
be subject to being adopted and approved on that ground as well.
For the reasons set forth herein, plaintiffs’ objection, [Doc. 1206], is OVERRULED and the
Magistrate Judge’s order, [Doc. 1191], is AFFIRMED AND ADPOTED by the Court and made the
ORDER of the Court.
So ordered.
ENTER:

s/J. RONNIE GREER
UNITED STATES DISTRICT JUDGE

27