EQUAL EMPLOYMENT OPPORTUNITY COMMISSION v. FEDEX GROUND PACKAGE SYSTEM, INC.

Filing 212

MEMORANDUM OPINION. Signed by Judge Mark R. Hornak on 3/21/18. (bdb)

IN THE UNITED ST ATES DISTRICT COURT
FOR THE WESTERN DISTRICT OF PENNSYLVANIA

EQUAL EMPLOYMENT
OPPORTUNITY COMMISSION,
Plaintiff,

)
)
)

)

Civil Action No. 2: 15-cv-256

)

v.

)

FEDEX GROUND PACKAGE
SYSTEM, INC.
Defendant.

Judge Mark R. Hornak

)
)
)
)
)

MEMORANDUM OPINION

The issue now before the Court centers on what, if any, discovery of electronically stored
information ("ESI"), specifically e-mails, should FedEx be required to produce in response to the
EEOC's discovery request. For the reasons that follow, FedEx's Motion to Preclude Discovery or
for Protective Order, ECF No. 118, is denied without prejudice.
This discovery dispute arises out of the EEOC's suit against FedEx in which the EEOC
alleges that FedEx engaged in nationwide and ongoing disability discrimination with respect to
deaf and hard-of-hearing Package Handlers and applicants to that position. 1
FedEx and the EEOC have met and conferred regarding e-mail production in this case. Email production, as it is relevant to the dispute here, has centered around e-mails that contain
specified search terms and are in the possession of FedEx's corporate custodians. The parties
engaged in back-and-forth deliberations on which corporate custodians and what search terms
would yield the appropriate discoverable e-mails. The Motion to Preclude indicates that the EEOC

1
This suit is a nationwide case in which the number of aggrieved individuals exceeds 300. The EEOC :;eeks
compensatory damages, punitive damages, and injunctive relief. See Comp!., ECF No. 1.

seeks to discover roughly 193 gigabytes of e-mail data (consisting of 363,765 e-mails and
attachments) in the possession of twenty-six corporate custodians. ECF No 118, ~ 1.
The core of FedEx's argument is that the search of its electronic repositories for the
requested "documents" would not tum up a sufficient quantity of responsive and relevant
documents to justify requiring FedEx to dig into 363,765 already-generated electronic files. In
addition, if the Court concludes e-mail discovery is relevant and proportional, FedEx argues theemails are "not reasonably accessible" within the meaning of Rule 26(b)(2)(B) because the volume
ofinformation2 exceeds FedEx's internal capabilities. Alternatively, if FedEx is to be required to
produce such ESI, FedEx requests that the EEOC bear the cost for all of that work.
A. The Statistical Sampling
Intending to demonstrate to the EEOC and the Court that the requested production was
beyond the scope of Rule 26, FedEx conducted a "statistical sample" of the e-mails. 3 ECF No.
119, at 6. FedEx took a sample size of 196,166 unique de-duplicated documents containing one or
more of the EEOC's search terms and reviewed a randomly selected group of 1,061 documents.

Id. at 6. FedEx initially asserted that of the 1,061 documents reviewed, only 111 (10.46%) were
responsive. Id. at 6-7. When FedEx included attachments and parent documents related to the

2

FedEx asserts that it has collected 2.13 8 terabytes (2, 13 8 gigabytes) of e-mail data, but the request at issue will
capture 193 gigabytes of that data. ECF No. 129, at 11. As a point ofreference, the current iPad Pros have a storage
capacity of either 64, 256, or 512 gigabytes. iPad Pro Technical Specifications, Apple, https://www.apple.com/ipadpro/specs/ (last visited March 20, 2018).
3
In its Brief in Support of its Motion to Preclude, FedEx references other cases that have used sampling to assess
large production requests. Of the three cases that FedEx cites, the relevant cases involve court-ordered sampling
after the court expressed initial doubts about a production request. FTC v. DIRECTV, Inc., No. 15-cv-01129, 2016
U.S. Dist. LEXIS 75480, at *4 (N.D. Cal. June 9, 2016) ("At this point, the relevance of these materials is largely
speculative and a random sample therefore gives DIRECTV what it seeks through these Requests for Production
while rendering the FTC's burden reasonably proportional to the materials' purported evidentiary value."); Vaigasi
v. Solow Mgmt. Corp., No. l l-cv-5088, 2016 U.S. Dist. LEXIS 18460, at *12 (S.D.N.Y. Feb. 16, 2016) (ordering a
sample search of a custodian's inbox after questioning a party's position with respect to producing e-mails); In re
Disposable Contact lens AntiTrust litig., No. 15-md-2626, 2016 U.S. Dist. LEXIS 151330, at *8 (M.D. Fla. Nov.
I, 2016) (resolving dispute about which sales manager custodians are subject to information searches by allowing
four custodians to be searched instead of eleven because it would provide a good sample of all eleven but not
addressing the type of statistical sampling at issue here).

2

reviewed e-mails (bringing the reviewed number to 2,816), only 170 were deemed responsive
(thereby dropping the "responsive rate" of the sample to 6%). Id. at 13. FedEx claimed that only
71 of those 170 documents would be discoverable due to assertions of privilege, and only 50 of
those 71 would not have been captured by other discovery requests not at issue here. 4 Finally,
FedEx asserts that those 50 documents do not contain any "smoking guns" that would support the
EEOC's claims. Id. at 119.

After the Court gave the EEOC the opportunity to "reverse

engineer" FedEx's statistical sample analysis, FedEx filed a notice with the Court "updating" its
prior numbers. ECF No. 184, at 1. Of the same 2,816 documents reviewed, FedEx now claims 235
should have been deemed responsive. This brings the responsive rate of the sampling to 8.3%.
FedEx claims that of that 235, only 79 documents potentially discoverable. Id. The EEOC objects
to many of these statistics. ECF No. 207, at 2. Specifically, the EEOC says FedEx is low-balling
the responsive rate since it points to over 200 documents that were marked by FedEx as
unresponsive that it believes should have been deemed responsive. Id.
FedEx seeks to weigh its response figures against its purported relevant costs. FedEx
estimates that the cost of running searches on the 193 gigabytes of e-mail data at issue will cost
$28,907 plus a $2,891 monthly hosting charge. 5 ECF No. 120, at 4.

B. Procedural History
The Court has read and considered the extensive briefing at ECF Nos. 119, 122, 124, 125,
129, and 130. The Court heard oral argument on the Motion to Preclude on June 22, 2017. ECF

4

FedEx reviewed those 170 documents and determined that 99 of them would be subject to the attorney-client
privilege or the work product doctrine and 21 of them have already been given to the EEOC or will be captured by
some other discovery request.
5 FedEx asserts that it has already paid over $73,000 to a third-party vendor to render e-mail data searchable. ECF
No. 120, ~ 13. Then in its Supplemental Brief, FedEx claims the EEOC's request here will cost over a million
dollars. ECF No. 208, at 1. FedEx adds its anticipated fees moving forward with this discovery request and the costs
for ingestion of data beyond the scope of this dispute (see footnote 9 infra) to reach its million-dollar mark. ECF No.
120, ~ 17. However, the costs FedEx will have to bear going forward for the discovery at issue here appears to be
the $28,907 figure and monthly $2,891 figure.

3

No. 138. After the oral argument, the Court deferred ruling until the parties exchanged information
related to FedEx's statistical input and methodology (used in FedEx's statistical analysis to support
its argument that there would be few fruits to its massive labor) in anticipation that the parties
could reach some sort of resolution. Order, ECF No. 141.
It appears that pretty much just the opposite occurred. The parties disagreed about the scope

of the Court's Order, requiring further direction from the Court. ECF Nos. 164, 165, and 166. The
parties then proposed a stipulated Order regarding such exchange of information as contemplated
by the Court's Order, ECF No. 170-1, and the Court approved it. ECF No. 172. Receiving no
supplemental briefings from the parties (with the exception of a Notice to the Court by FedEx on
December 19, 2017, ECF No. 184, correcting various statistical figures reported in its initial
briefings), the Court ordered the parties to submit a joint status report regarding whether there was
still a live dispute as to this matter. Order Directing Joint Status Report, ECF No. 190. The parties
indicated that they wanted to submit supplemental briefs on that issue, and those briefs have been
submitted to and considered by the Court. ECF Nos. 194, 207, 208, and 209.

C. Discussion
After its review of the extensive materials in what appears to the Court to be a relatively
straightforward discovery dispute, the Court concludes that the EEOC's requested ESI from FedEx
fits within the limits and scope of discovery that the Federal Rules of Civil Procedure say should
be permitted in this case. FedEx reportedly went to great lengths (and expense) to demonstrate that
the EEOC's requests were "unreasonable," based on the magnitude of the search results they
yielded, but even when the Court considers FedEx's own statistics-based arguments, it concludes
that the EEOC's requested discovery remains "relevant to [its] claim or defense and proportional
to the needs of the case, considering the importance of the issues at stake in the action, the amount

4

in controversy, the parties' relative access to relevant information, the parties' resources, the
importance of the discovery in resolving the issues, and whether the burden or expense of the
proposed discovery outweighs its likely benefit." Fed. R. Civ. P. 26(b)(l).
FedEx argues that regardless of accessibility and the burden of producing such discovery,
both the search terms themselves and the results produced in the sample analysis (and therefore
the results that would be produced in a full-scale analysis) are irrelevant to the EEOC's claims in
this case, and this alone justifies granting the Motion to Preclude under Rule 26. The Court agrees
with the EEOC that FedEx incorrectly attempts to frame relevancy as a binary test where
information is either irrelevant or a "smoking gun." After review of the EEOC's e-mail discovery
request, it strikes the Court as within the much more encompassing target of what would
reasonably lead to the discovery of information relevant to the EEOC's claims or FedEx's
defenses.
The Court next turns to FedEx's accessibility and undue burden arguments. FedEx claims
that the total number of documents that would be subject to the EEOC's e-mail production request
is in the ballpark of 363,765. Taking into account FedEx's revised responsive rate based on its
own sample analysis of 8.3%, the EEOC's request is likely to generate somewhere around 30,192
responsive documents. 6 If the EEOC's assessment that there were an additional 200 responsive

documents in the sample batch is correct, the response rate then goes to 15% 7 or about 54,565 out

6

The record does not reflect that FedEx proposed any further meaningful revision of the search terms teed up via the
EEOC that would both address the core discovery legitimately sought by the EEOC but would also screen out what
FedEx says may be well more than 200,000 false hits. It seems to the Court that the able lawyers on both sides of the
case could figure out what it is about the last-used search terms that is driving that result. That appears to be the best
(and most correct) solution to the current dispute. In these regards, FedEx has presented a legitimate argument that
further refinement of search terms makes a lot of good sense. Where FedEx's argument goes offtrack is its position
that no further refinement should occur and ES! searching and production should end, and in essence, the whole thing
should be called off. ECF No. 119, at 7.
7
Those extra 200 documents added to FedEx's 235 responsive documents equals 435. That number divided by the
total pool (2,816) is 0.1544 or 15%.

5

of 363,765 such documents. ECF Nos. 207, at 2; 209, at 6. FedEx relies on the argument that
because the time and expense of conducting searches through nearly two hundred thousand
electronic documents is so burdensome, the EEOC should be precluded from obtaining what will
likely be over 30,000 (or possibly 54,000+) properly discoverable documents (minus any
legitimately withheld by privilege).
The Court does not accept the argument that because the starting pool of documents is that
large, and the anticipated relevant results are comparatively smaller (although still well into the
tens of thousands), an opposing party should not be able to access via discovery any discoverable
ESI documents. Many organizations, and indeed even some individuals, choose to store really
large amounts of data electronically, presumably because they have concluded that doing so
furthers their important day-to-day interests and, on balance, is the best/most efficient method for
compiling and storing that information. The reality of "e-discovery" is that parties are left with the
tasks of examining and then as applicable producing their electronic information so compiled and
stored when called upon in litigation. But the fact that these repositories create complex
mechanisms to store huge amounts of information cannot be used in and of itself as a shield to
avoid discovery requests otherwise permitted under the Federal Rules of Civil Procedure. 8
After the parties were able to reverse engineer FedEx's self-constructed sample analysis,
the EEOC maintained its position that its "e-discovery" request does not run afoul of the applicable
Rules of Civil Procedure, and the Court generally agrees. Putting aside the dispute as to whether
FedEx initially "understated" the extent to which various searches produced responsive
documents, the production request is facially within the scope of permitted discovery, and the

8

Nor does the Court conclude that it is relevant for these purposes that many of the searched documents had been
retained by FedEx due to a preservation order in another lawsuit. ECF No. 119, at 2. The bottom line is that the subject
of the requests is accessible. That may very well be a lucky break for the EEOC, but luck (or simply the reason
information is accessible) is not a factor contemplated under Rule 26.

6

production requested strikes the Court as being proportional to the needs of the case. The parties
should also bear in mind that, at the beginning, they jointly requested nearly four (4) years of fact
discovery. Federal Rule 26(f) Report of the Parties, ECF No. 89, at 3. Thus, it is beyond doubt that
both parties clearly understood, from the beginning, that this was going to be a big case that would
take some considerable time and effort to unwrap and prosecute/defend, and they argued just that
to the Court. It also strikes the Court that perhaps had FedEx initially focused its considerable legal
energy on working with the EEOC to conduct the sample analysis and then refine (and perhaps
refine again, and if necessary, again) search terms, FedEx might have saved time, resources, and
expense by avoiding work that was performed to generate and support an argument, as opposed to
just doing that discovery work. Now this issue has been stalled so that the EEOC could be given
the opportunity to reverse engineer FedEx's results in order to test and then properly respond to
FedEx's legal arguments. The Court does find it telling that after this "collaboration" occurred, the
EEOC's arguments remained largely unchanged while FedEx's "results" from the analysis did
notably change.
As the Court has noted above, from "Day One," the parties have been of the view that they
believe this is a big case. On that, the Court takes them at their word. But, part and parcel of that
reality is that in big cases, there will likely be large amounts of potentially d[scoverable
information, which means, for better or worse, a lot of work in identifying and producing it. At the
end of that process, there still will (not surprisingly) be a lot of information that will become the
grist of the litigation mill. That is apparently the case here.

D. Conclusions
Considered as a whole, at the end of the day, the Court concludes that FedEx has not carried
its burden under Rule 26(b)(2)(B) to show that the information is not reasonably accessible

7

because of undue burden or cost (it is accessible, because it has in fact been accessed), or that the
costs of sifting through the repositories that FedEx has elected to maintain as part of its general
business plans and operations should be shifted to the EEOC. In reaching these conclusions, the
Court has considered the underlying discovery requests, the responses of the parties, the results of
the original and subsequent analyses offered by FedEx and the EEOC, and the parties' respective
positions to those analyses, all in light of the actual claims asserted by the EEOC in this litigation,
and the various, and at times fluid, defenses advanced by FedEx. The Court can and does
confidently come to the following conclusions:
•

The scope, nature, and content of the EEOC's discovery requests (certainly as to the
"HQ custodians" now at the center of this dispute) as they sweep in ESI of FedEx are
logical, facially germane, reasonably calculated to lead to the discovery of relevant
information, and are proportional to the nature and needs of this case;

•

The level of search result responsiveness, even considered through the relevancedirected lens seemingly applied by FedEx, is not so low, scant, limited or undue, no
matter whose various analytical/argumentative numbers are used, to put the brakes on
the EEOC's discovery request in toto, and in any event what it yields is a meaningful
amount of discoverable documents (at least 30,000+);

•

FedEx's position that all discovery of ESI should be denied or cut off on the basis that
in its estimation, there is not enough of it generated so as to make it valuable, useful,
relevant (or otherwise necessary at all) has no substantive or substantial basis in the
Rules, in the case law, or in logic, "all or nothing" is not a principle of foderal civil
discovery law. In the Court's estimation, the EEOC has demonstrated that there are
more than enough relevantly responsive "hits" from the latest round of more fine-tuned
ESI searches to convince the Court that FedEx's Motion to essentially bar any
discovery of ESI has no merit in that regard;

•

Given the contours of Fed. R. Evid. 502(d), any actual burden on FedEx to search the
identified and segregated ESI items can be greatly reduced, given that the EEOC has
agreed to take over that task at its expense in the context of the "clawback" provisions
of Rule 502(d) and any applicable agreement of the parties in such regards;

•

Even if FedEx is concerned that there is a privilege issue in the EEOC doing the
examination of the documents (even within the protection of Fed. R. Evid. 502(d) ), it
appears to the Court that there are reasonable, targeted, limited custodian sc::~arches that
are highly likely to ameliorate that privilege risk;

8

•

That for all of the respective huffing and puffing that is seemingly part of civil
discovery disputes in this day and age, it firmly appears to the Court that in large
measure, the dispute that now exists and which is central to this Motion stems mostly
from FedEx's desire to respond to the ESI component of the EEOC discovery request
on the terms as it defines them. For better or worse, while our system of civil discovery
relies on the parties and their lawyers to make the first decisions about the production
of documents (electronic or otherwise) within the contours of the applicable Rules, they
don't get to define those Rules;

•

And, as to the issue of costs/expenses, the Court concludes that the EEOC' s willingness
to do the leg work, within the protections of Fed. R. Evid. 502( d), addresses any such
cost/expense issues.

The Court does not anticipate that it will analyze individual e-mails with respect to this
dispute or any related disputes that may arise, and while in theory it may have to appoint a Special
Master to do so if it comes to that, the Court does not at the moment anticipate that such would be
necessary, since this really is not a complicated situation. And, beyond that, the costs associated
with further refinement and production from the currently identified document pool do not strike
the Court as being disproportionate to this case and the issues in it. The parties can continue to
meet and confer to work out ESI issues as they come up, including further collaboration on refining
the search terms at issue in this dispute. The Court trusts that the parties and their well-seasoned
counsel, with this ruling in hand, can move this case forward as to this issue. 9 The parties shall
bear their own fees and costs associated with this Motion, and shall file a joint status report on or
before April 23, 2018, as to the status of these matters.
An appropriate Order will issue.

Mark R. Hornak
United States District Judge
cc: All counsel ofrecord
9

This Memorandum Opinion and the accompanying Order does not specifically resolve FedEx's "forward-looking"
request for relief as to potential searches into ES! related to "field custodians." But, from the Court's perspective, that
request for reliefappears to be premature, and the parade of horribles rolled out by FedEx seems to be quite speculative
at this juncture. And, as the Court noted above, we are where we are in large measure due to FedEx deciding that this
process had gone far enough, and thereby cutting off the further reduction of the document pool that would likely
come from further refinement of search terms.

9