Proportionality, Pretrial Confidentiality, and Discovery Sharing
Proportionality, Pretrial Confide ntiality, and Discover y Sharing
Dustin B. Benham 0 1 2
0 Thi s Article is brought to you for free and open access by the Law School Journals at Washington & Lee University School of Law Scholarly Commons. It has been accepted for inclusion in Washington and Lee Law Review by an authorized administrator of Washington & Lee University School of Law Scholarly Commons. For more information , please contact , USA
1 Dustin B. Benham, Proportionality, Pretrial Confidentiality, and Discovery Sharing, 71 Wash. & Lee L
2 Texas Tech University School of Law , USA
Confidentiality, and Discovery Sharing
Dustin B. Benham
I. Introduction ...................................................................2182
1. Upfront Sharing Provisions...............................2202
A. History .....................................................................2214
B. Current Proposal .....................................................2219
IV. A Framework for Discovery Sharing.............................2225
A. Upfront Sharing Provisions Versus Later
Associate Professor of Law, Texas Tech University School of Law. I
would like to thank Richard Marcus for his thoughtful comments on an earlier
draft of this Article. Thanks also to Meryl Benham, Bryan Camp, Elizabeth
Caulfield, Rory Ryan, and Natalie Tarenko for their contributions to this work. I
am especially thankful to my research assistant Nicole Sears, J.D. Candidate
2015, Texas Tech University School of Law, for her excellent research and
Courts around the country enter pretrial confidentiality
orders every day.1 Commentators and courts have long debated
the impact of court confidentiality on public safety.2 This debate
often centers on whether public harms flow from court orders
that limit the audience for discovery information that those
same courts order parties to produce.3 But an underexplored
aspect of pretrial confidentiality is the cost it imposes on the
litigation system. Confidentiality orders that prevent litigants
in similar cases from sharing information make litigation less
efficient and less effective.4 Courts are split on if, when, and how
1. See FED. R. CIV. P. 26(c)(1)(g) (allowing protective orders to limit access
to trade secrets and other proprietary information); see also, e.g., Laurie Kratky
Dore, Secrecy by Consent: The Use of Limits of Confidentiality in Pursuit of
Settlement, 74 NOTRE DAME L. REV. 283, 285 (1999) (noting the frequent use of
confidentiality in civil litigation).
2. Compare Richard L. Marcus, Myth and Reality in Protective Order
Litigation, 69 CORNELL L. REV. 1, 6–9 (1983) (chronicling the intrusive nature of
civil discovery and observing that parties have legitimate reasons to keep
discovery information private), and Richard P. Campbell, The Protective Order
in Products Liability Litigation: Safeguard or Misnomer?, 31 B.C. L. REV. 771,
772–75 (1990) (contending that privacy interests should trump public and party
interests broader dissemination of pretrial discovery), with Dore, supra note 1,
at 296 (noting the view that courts are “publicly funded” and “accountable to
and guardians of a broader public interest”), Richard Zitrin, The Judicial
Function: Justice Between the Parties, or a Broader Public Interest?, 32 HOFSTRA
L. REV. 1565, 1567 (2004) (contending that private dispute resolution is just one
function of courts; they also have a significant public function), Joseph F.
Anderson, Jr., Secrecy in the Courts: At the Tipping Point?, 53 VILL. L. REV. 811,
813 (2008) (noting that “[t]he debate over ‘court-ordered’ secrecy has festered for
several decades” and concluding that more court transparency is needed), and
Sunshine in Litigation Act of 2009: Hearing on H.R. 1508 Before the Subcomm.
on Commercial & Admin. Law Comm. on the Judiciary, 111th Cong. 1 (2009)
(testimony of Leslie A. Bailey, Staff Att’y, Public Justice) (“In short, through
protective orders, secret settlements, and sealed court records, the public courts
are being used by private parties to hide smoking-gun evidence of wrongdoing.”).
3. See, e.g., Anderson, supra note 2, at 811–12 (recounting a fatal accident
stemming from allegedly faulty tires where information about the defect,
potentially available from over two hundred related lawsuits, was kept from
victims by confidentiality orders).
4. See, e.g., Ward v. Ford Motor Co., 93 F.R.D. 579, 580 (D. Colo. 1982)
(determining that to deny sharing between litigants “would be tantamount to
holding that each litigant who wishes to ride a taxi to court must undertake the
expense of inventing the wheel”).
these orders should allow litigants to share pretrial discovery
This Article examines the systemic harms of confidentiality
orders and contends that courts should exercise their discretion
to allow information sharing between similar cases absent
substantial evidence of countervailing factors. It is the first
significant piece of legal scholarship to examine court
confidentiality and discovery sharing in light of the recent and
contentious proposed amendments to the Federal Rules of Civil
Procedure, approved in May 2014 by the Federal Judicial
Conference Standing Committee.6
The current General Motors defective-ignition crisis provides
a contemporary example of the potential systemic impact of
restrictive confidentiality orders. While many rightly focus on the
potential impact of court confidentiality on public safety,7 court
orders also likely prevented some of the plaintiffs in virtually
identical lawsuits from sharing discovery information with each
other. By “silo-ing” discovery information in this way, G.M. (like
many other litigants engaged in national-scale litigation) forced
5. Compare Long v. TRW Vehicle Safety Sys., Inc., No.
CV-09-2209-PHXDGC, 2010 WL 1740831, at *1 (D. Ariz. Apr. 29, 2010) (denying request for
sharing provision in protective order out of preference for later intervention and
modification), and Wal-Mart Stores E., L.P. v. Endicott, 81 So. 3d 486, 490 (Fla.
Dist. Ct. App. 2011) (denying request for discovery-sharing provision based on
conflict with state trade-secret statute), with Garcia v. Peeples, 734 S.W.2d 343,
346 (Tex. 1987) (mandating discovery-sharing provision in protective order
absent showing of prejudice to producing party), and Raymond Handling
Concepts Corp. v. Super. Ct., 45 Cal. Rptr. 2d 885, 888 (Cal. Ct. App. 1995)
(allowing trial court discretion to include discovery-sharing provision).
6. See ADVISORY COMM. ON CIVIL RULES, REPORT TO THE STANDING
COMMITTEE 91–94 (May 2, 2014), http://www.uscourts.gov/uscourts/RulesAnd
Policies/rules/Agenda%20Books/Standing/ST2014-05.pdf [hereinafter COMMITTEE
REPORT] (enumerating proposed amendments to the Federal Rules of Civil
Procedure); Tera E. Brostoff, Changes to E-Discovery Approved; “It Should Work,”
Professor Predicts, 82 U.S.L.W. 1918 (June 2, 2014) (discussing the rationale
behind the amendments to Rule 37 and predicting the effectiveness of the new
7. See, e.g., Bill Vlasic, Inquiry by G.M. Is Said to Focus on Its Lawyers,
N.Y. TIMES, May 18, 2014, at A1 (“G.M.’s unwillingness to share information it
had about defective switches with regulators most likely cost lives in
its opponents to wastefully reinvent the wheel when conducting
discovery in each case.8
Indeed, plaintiffs with identical discovery requests in two
virtually identical cases might be forced to litigate their
entitlement to the same information anew in both cases. If courts
in similar cases simply allowed parties to share discovery, the
twin lawsuits would consume fewer court and party resources.9
Multiplied over many cases and many document requests, court
orders that prevent sharing create a significant drag on the
system. Based on these efficiency concerns, most commentators
that have considered the issue support some form of discovery
Surprisingly, however, courts are still split on the propriety
of protective-order provisions that allow information sharing
between similar cases.11 Some courts have mandated discovery
sharing, holding that courts abuse their discretion by issuing
orders that prevent the practice.12 At the other end of the
spectrum, some courts have recently held that discovery-sharing
provisions are forbidden in many cases.13 Indeed, federal and
8. Cf., e.g., Garcia, 734 S.W.2d at 346–48 (finding that nonsharing
protective orders, in the products liability context, create inefficiency by forcing
parties to conduct discovery without the benefit of information from previous
9. See, e.g., id. at 347 (“In addition to making discovery more truthful,
shared discovery makes the system itself more efficient.”).
10. See, e.g., Richard L. Marcus, The Discovery Confidentiality Controversy,
1991 U. ILL. L. REV. 457, 464–66, 466 n.61, 495–96 (cataloguing pro-sharing
action by the legislatures and an ABA commission while endorsing discovery
sharing); Arthur R. Miller, Confidentiality, Protective Orders, and Public Access
to the Courts, 105 HARV. L. REV. 427, 498–99 (1991) (favoring discovery sharing
when it actually promotes efficiency); Dore, supra note 1, at 363–65 (same).
11. See, e.g., cases cited supra note 5 (listing numerous cases in which
courts have come to differing conclusions concerning protective orders).
12. See Garcia v. Peeples, 734 S.W.2d 343, 346–48 (Tex. 1987) (“The facts of
this case do not justify the blanket protective order, and in rendering an
overbroad order, the trial court abused its discretion.”).
13. See Wal-Mart Stores E., L.P. v. Endicott, 81 So. 3d 486, 490 (Fla. Dist.
Ct. App. 2011) (denying a motion for a protective order on the grounds that it
would allow dissemination of trade secrets in violation of a state statute); cf.
Long v. TRW Vehicle Safety Sys., Inc., No. CV-09-2209-PHX-DGC, 2010 WL
1740831, at *1 (D. Ariz. Apr. 29, 2010) (finding that the presumption of public
access “is trumped where the defendant establishes good cause to protect
state courts across the country, applying similar protective-order
rules, are increasingly fractured on the issue.
The dispute about discovery sharing is particularly curious in
the face of recent and ongoing procedural reform designed to
make pretrial litigation more efficient. In late May 2014, the
Federal Judicial Conference Standing Committee on the Rules of
Practice and Procedure approved significant revisions to Federal
Rule of Civil Procedure 26 to reduce the systemic burdens of
pretrial discovery.14 These proposed amendments are almost
certain to become law in December 2015.
The centerpiece of the changes would highlight the concept of
“proportionality” as the scope of discovery.15 Proportionality is the
notion that resources expended on discovery in a given case
should correlate with the benefits of conducting the discovery.16
discovery sharing.17 But sharing works in tandem with
proportionality by allowing parties to more accurately tailor
discovery in cases where the parties are informed by shared
information. Indeed, the crux of the amendments’ letter and
spirit—minimizing the undue burdens of pretrial discovery18—
14. See COMMITTEE REPORT, supra note 6, at 63–72 (detailing proposed
amendments to the Federal Rules of Civil Procedure); Brostoff, supra note 6
(describing the rationale behind adopting certain amendments to the rules:
“focusing on the goals of cooperation, proportionality and early case
15. See COMMITTEE REPORT, supra note 6, at 65 (incorporating
proportionality as Rule 26(b)(1) scope of discovery).
16. See FED. R. CIV. P. 26(b)(2)(C)(iii) (limiting discovery when “the burden
or expense of the proposed discovery outweighs its likely benefit, considering the
needs of the case, the amount in controversy, the parties’ resources, the
importance of the issues at stake in the action, and the importance of the
discovery in resolving the issues”); see also 8A CHARLES ALAN WRIGHT ET AL.,
FEDERAL PRACTICE AND PROCEDURE § 2036 (3d ed. 2010) (discussing the rationale
behind Rule 26(b)(2)(C)’s prohibition on “disproportionate discovery demands”);
Richard Marcus, Procedural Postcard from America, 1 RUSS. L.J. 9, 19–20 (2013)
(comparing American notions of proportional discovery with European norms).
18. See id. at 82 (“The objective is to guard against redundant or
disproportionate discovery by giving the court authority to reduce the amount of
discovery that may be directed to matters that are otherwise proper subjects of
stands in stark contrast to decisions that require parties to craft
discovery requests in the dark and re-litigate the same discovery
issues in virtually identical cases.
Beyond the extent to which discovery sharing complements
recent procedural reform efforts, it also has many collateral
benefits, including increased litigation integrity through
transparency along with, in at least some cases, increased
accountability for wrongdoing.19
Part II of this Article examines the court-confidentiality
problem and its relationship to discovery sharing. Part III
examines historical and contemporary trends in pretrial
discovery and their relationship to court confidentiality and
discovery sharing. Then, in Part IV, the Article proposes some
baseline discovery-sharing principles and examines the potential
interaction between these principles and the current proposed
amendments to the civil rules.
II. Court Confidentiality
Popular wisdom holds that courts are open to the public. As
Atticus Finch presented his unforgettable closing argument in To
Kill a Mockingbird, spectators watched from packed galleries.20
Though unjustly divided by race between the lower gallery and
the balcony, they observed both the presentation of evidence in
the trial and the unjust conviction of Atticus’s client.21
19. See, e.g., Garcia v. Peeples, 734 S.W.2d 343, 347 (Tex. 1987) (“Shared
discovery is an effective means to insure full and fair disclosure. Parties subject
to a number of suits concerning the same subject matter are forced to be
consistent in their responses by the knowledge that their opponents can
compare those responses.”); Dore, supra note 1, at 363–65 (recognizing implicitly
that shared disclosure may increase accountability by noting that “[t]he desire
to shield oneself from other potential claims, however, does not alone justify”
objection to discovery sharing).
20. See HARPER LEE, TO KILL A MOCKINGBIRD 186–87, 229–35 (Harper 2002)
(describing the overflowing public courtroom, indicating that not even standing
room remained to watch the case unfold); see also TO KILL A MOCKINGBIRD
(Universal Studios 1962) (using the packed courtroom, open to the public, to
demonstrate that the case had grasped the attention of the public).
21. LEE, supra note 20, at 240.
Popular notions of openness aside, the idea that courts and
their processes are open to the public is only partially accurate.
While a modern public would no doubt have a constitutional right
of access to a criminal or civil trial, much of what takes place in
litigation is not open to public view. In particular, civil discovery
is often kept confidential through a procedural device known as a
protective order.22 And because many cases settle during
discovery and before trial, the general public and other litigants
involved in virtually identical cases have limited access to
pretrial discovery information.23
Confidentiality in litigation, however, comes at a cost.
According to pro-transparency advocates, keeping public health
and safety hazards confidential costs lives.24 Beyond the public
safety issue, litigating individual cases in isolation has a systemic
impact. Indeed, some have argued that the court system is made
more expensive and less effective when those same pretrial
confidentiality orders forbid litigants from sharing discovery in
one case with litigants in other virtually identical cases.25
22. See FED. R. CIV. P. 26(c)(1)(g) (permitting a court to issue a protective
order to ensure that “a trade secret or other confidential research, development,
or commercial information not be revealed or be revealed only in a specified
way”); see also Judith Resnick, The Privatization of Process: Requiem for and
Celebration of the Federal Rules of Civil Procedure at 75, 162 PENN. L. REV.
1793, 1814–17 (2014) (describing declining public access to once-public court
functions, including pretrial discovery).
23. See, e.g., Andrew D. Goldstein, Sealing and Revealing: Rethinking the
Rules Governing Public Access to Information Generated Through Litigation, 81
CHI.-KENT L. REV. 375, 401–02 (2006) (noting that much “litigant-centered”
information, like discovery, generated through court processes is often kept from
the public); see also Richard A. Zitrin, The Laudable South Carolina Rules Must
Be Broadened, 55 S.C. L. REV. 883, 886–87 (2004) (describing the lack of judicial
awareness of lawyer-driven secrecy through discovery and settlements occurring
outside the view of the court).
24. See Zitrin, supra note 2, at 1565–66 (describing a particular case in
which a plaintiff’s attorney, after agreeing to a “secretized” settlement,
“[acknowledged] that others may have died later as a consequence” of the
25. Cf. Dore, supra note 1, at 363 (“Discovery sharing, while arguably
undermining the efficiency of discovery in the immediate lawsuit, potentially
avoids the wasteful duplication of discovery in collateral litigation, thereby
ultimately advancing the efficient resolution of disputes.”).
As a prefatory matter, I choose the word “confidential”
carefully here. Terminology in the confidentiality debate can
carry a loaded meaning.26 Proponents of more confidentiality, for
instance, often refer to the issue as one of “privacy.”27 Likewise,
those who favor more public access to discovery materials
typically refer to the issue in terms of “secrecy.”28 In this Article, I
refer to those in favor of less public access as “pro-confidentiality”
and those in favor of more public access as “pro-transparency,”
recognizing that neither term connotes those positions with
The following subparts explore the basic protective-order
framework in the American system and narrower, yet significant,
problems stemming from the use of pretrial protective orders to
stifle discovery sharing.
A. Pretrial Protective Orders
For more than three decades, courts, scholars, and legislators
have debated the impact of court confidentiality. For the most
part, the discussion has centered on public access and the tension
between the idea that courts are public institutions that express
public values and the notion that litigants do not sacrifice all
privacy at the courthouse door.30 And while it is difficult to
calculate precisely which side has the momentum, it is
undeniable that pretrial confidentiality is a feature of modern
26. Cf. Nancy S. Marder, Introduction to Secrecy in Litigation, 81
CHI.KENT L. REV. 305, 305 (2006) (suggesting a court-confidentiality vocabulary).
27. See, e.g., Marcus, supra note 2, at 15 (observing the “privacy afforded
pretrial proceedings” through protective orders). But cf. Marcus, supra note 10,
at 457 (referring to the issue as “The Discovery Confidentiality Controversy”).
28. See, e.g., Zitrin, supra note 2, at 1572 (“In the last five years, secrecy in
settlements has become an increasingly common subject of articles in the
popular legal press and more scholarly forums.”).
29. See Marder, supra note 26, at 305 (describing the connotations
associated with certain “loaded terms” often used when addressing the issue of
confidentiality in discovery).
30. See sources cited supra note 2 (listing works of scholarship
exemplifying the debate surrounding privacy in discovery).
31. See, e.g., Dore, supra note 1, at 285 (observing that secrecy is a part of
The confidentiality problem is typically framed as pitting the
public’s interest in knowing against the litigant’s interest in
confidentiality. In reality, this false binary oversimplifies the
problem and ignores a third category affected by court
confidentiality—litigants in separate but similar cases.32 This
group occupies a special position, with distinct interests in
accessing discovery information but still regularly excluded from
access by confidentiality orders.33
The trend of keeping pretrial litigation information
confidential from both the public at large and other similar
litigants pervades all aspects of the trial process. Before trial,
courts and attorneys working on behalf of litigants liberally
employ confidentiality orders to closet pretrial discovery from
anyone outside of each particular case.34
Many of these cases settle, and settlement agreements often
contain court-enforceable gag provisions that extend the original
confidentiality orders indefinitely.35 These agreements, like the
orders that precede them, often include “return-or-destroy”
provisions that require the parties to return or destroy all
discovery information in the case within a few months of
settlement.36 These mechanisms make it more difficult for the
every phase of civil litigation).
32. See, e.g., Marcus, supra note 10, at 493–94 (discussing “access for other
33. Cf., e.g., id. (summarizing steps being taken to improve access to
confidential material produced during discovery for similarly situated litigants).
34. See, e.g., Dore, supra note 1, at 332 (“To expedite discovery and avoid
repeated motions for a protective order regarding every document believed to be
confidential, parties will frequently agree to, and courts will regularly issue,
umbrella protective orders.”).
35. See, e.g., id. at 290–92 (cataloguing the Civil Rules explicit bent toward
settling cases); Anderson, supra note 2, at 813 (describing, from a federal judge’s
perspective, the “take it or leave it” approach of some parties in getting courts to
ratify secret settlements through court order).
36. See, e.g., Anderson, supra note 2, at 814 (describing requests from one
or both of the parties to enter orders with “return or destroy” provisions). Even
if relevant material is returned to the producing party, that party very well may
have a duty to preserve it for future litigation. See, e.g., 8A WRIGHT ET AL., supra
note 16, § 2284.1 n.4 (discussing the possible sources of an obligation to preserve
information within control of a party). Compliance with that duty, however, is
made more difficult to police when protective orders force the return or
destruction of the material by opposing parties.
general public and litigants in similar cases to obtain the
information. Even in cases that make it to trial, some courts have
recently, and inexplicably, sealed trial testimony and exhibits
from public view.37
Almost all facets of the broader court-confidentiality problem
stem from a common root—protective orders. Confidential
discovery, via protective orders, takes place in many cases that
are ultimately settled.38 If parties were unable to keep
information exchanged in discovery secret in the first place,
secret settlements would be largely ineffective.39 As a result, it is
not surprising that the early confidentiality controversy centered
on appropriateness of protective orders.
The term “protective order” actually describes several related
procedural devices that limit discovery.40 But the type of
protective order most relevant to the court-confidentiality debate
allows courts to limit the persons to whom discovery information
may be disclosed.41 To guarantee the confidentiality of pretrial
discovery materials, these orders are necessary because parties
may freely disseminate anything they learn in discovery to the
public at large, absent a contrary court order.
Described in Rule 26(c)(1)(g), these audience-limiting orders
mandate that “a trade secret or other confidential research,
37. See, e.g., Idar v. Cooper Tire & Rubber Co., No. C-10-217, 2011 U.S.
Dist. LEXIS 77568, at *4 (S.D. Tex. July 18, 2011) (permitting third-party
intervenors to obtain documents produced by defendant in the current case,
conditioned upon signing the parties’ confidentiality agreement, and denying
access to exhibits produced by defendant in previous case under a different
38. See Dore, supra note 1, at 384–85 (describing the relationship between
confidentiality and settlement).
39. See id. at 384 (“Secrecy undoubtedly facilitates the settlement process,
and in some cases, compromise could not be reached without some assurance of
40. See FED. R. CIV. P. 26(c)(1)(a)–(g) (enumerating various examples of
“good cause” for a court to grant a protective order); see also 8A WRIGHT ET AL.,
supra note 16, § 2036 (stating that “a court is not limited to the eight specified
types of orders” enumerated in Rule 26(c)).
41. See FED. R. CIV. P. 26(c)(1)(g) (allowing court use of a protective order to
prevent sensitive commercial information from being disseminated to
competitors); see also FED. R. CIV. P. 26(c)(1)(f) (allowing courts to seal
development, or commercial information not be revealed or be
revealed only in a specified way.”42 A particular order under this
provision might limit disclosure of discovery materials to just
those participating in the litigation—the parties, their attorneys,
support staff, and experts.43 A more onerous, and rare, version
restricts access to only attorneys and consultants—precluding
even parties from having access to particularly sensitive
The defining feature of an audience-limiting protective order
is that it allows discovery to go forward while simultaneously
limiting the audience for the materials. In a way, this type of
order might be viewed as a pragmatic middle ground.45
Implicit in the protective-order arrangement is the notion
that the primary purpose of discovery is to prepare and resolve
litigation.46 Protective orders are often geared to steer discovery
information to the appropriate audience for this purpose, even at
the expense of other legitimate audiences.47
Nevertheless, under Rule 26, confidentiality is not automatic
or an entitlement. Indeed, like other protective orders, Rule
26(c)(1)(g) confidentiality orders may be entered only on a
showing of “good cause.”48 Under the best formulations of good
42. FED. R. CIV. P. 26(c)(1)(g).
43. See Dore, supra note 1, at 327 (“Courts, for example, may restrict the
disclosure of such discovery to designated persons or forbid its use for purposes
unrelated to the preparation and settlement of the case at hand.”).
45. Cf. Marcus, supra note 10, at 484–85 (observing that a broad right of
public access to discovery information would disrupt discovery).
46. Cf. Marcus, supra note 2, at 7 (contending that those involved in the
litigation system operate under “the assumption that any use of discovery
materials except to prepare for trial is inappropriate”). But see Dustin B.
Benham, Dirty Secrets: The First Amendment in Protective-Order Litigation, 35
CARDOZO L. REV. 1781, 1808 (2014) (acknowledging that the primary purpose of
discovery is to prepare for trial but observing that “[e]ven in private litigation,
parties openly use the system to obtain information” among other collateral
47. See Benham, supra note 46, at 1806–08 (arguing that the dissemination
of discovery information is protected expression).
48. See, e.g., Pia v. Supernova Media, Inc., 275 F.R.D. 559, 560 (D. Utah
2011) (“Under the rule, the party seeking protection has the burden to show
cause, a party seeking a protective order must make a
particularized showing of both the confidential nature of the
information and the harm that would flow from publicly
disclosing it.49 Conclusory allegations, unsupported by specific
evidence, will not justify an order.50
Still, trial courts have inarguably broad discretion in finding
good cause and fashioning protective orders, “frequently finding
protection justified, and frequently denying protection.”51 Even
discretion, however, has its limits. Multiple courts have
articulated good cause factors to provide at least a minimal
But in practice, the good cause standard has proven to be
largely friendly to routine confidentiality requests, allowing many
good cause for preventing dissemination of discovery materials.”).
49. See, e.g., Gelb v. Am. Tel. & Tel. Co., 813 F. Supp. 1022, 1034 (S.D.N.Y.
1993) (“With respect to [a] claim of confidential business information, [the good
cause] standard demands that the company prove that disclosure will result in a
clearly defined and very serious injury to its business.”); Parsons v. Gen. Motors
Corp., 85 F.R.D. 724, 726 (N.D. Ga. 1980) (finding the “good cause” requirement
“to mean that the party seeking the protective order must demonstrate that the
material sought to be protected is confidential and that disclosure will create a
competitive disadvantage for the party”).
50. See, e.g., Glenmede Trust Co. v. Thompson, 56 F.3d 476, 483 (3d Cir.
1995) (“‘Good cause’ is established when it is specifically demonstrated that
disclosure will cause a clearly defined and serious injury. Broad allegations of
harm, unsubstantiated by specific examples, however, will not suffice.”); cf. 8A
WRIGHT ET AL., supra note 16, § 2035 (“The courts have insisted on a particular
and specific demonstration of fact, as distinguished from stereotyped and
conclusory statements, in order to establish good cause.”); Richard L. Marcus, A
Modest Proposal: Recognizing (at Last) that the Federal Rules Do Not Declare
that Discovery Is Presumptively Public, 81 CHI.-KENT L. REV. 331, 348 (2006)
(noting that the standards for issuing protective orders are “somewhat exacting”
but also contending that the substantial burden for obtaining a protective order
does not imply public right of access in the absence of one).
52. See, e.g., In re Roman Catholic Archbishop, 661 F.3d 417, 424 (9th Cir.
2011) (balancing the factors enumerated in Glenmede Trust Co. to determine
whether a party has shown “good cause”); Glenmede Trust Co., 56 F.3d at 483
(enumerating several factors, “which are neither mandatory nor exhaustive,
that may be considered in evaluating whether ‘good cause’ exists”); Mosaid
Techs., Inc. v. LSI Corp., 878 F. Supp. 2d 503, 508 (D. Del. 2012) (“Assessing
whether good cause exists to seal a judicial transcript generally involves a
balancing process, in which courts weigh the harm of disclosing information
against the importance of disclosure to the public.”).
courts to enter protective orders as a matter of course in complex
cases.53 Other courts have recently gone as far as creating a
standing protective-order form to which the parties may
stipulate, standardizing the rubber-stamp process.54
B. Fault Lines in the Protective-Order Debate
In response to a system that makes it easy to conceal
discovery information, pro-transparency advocates have waged a
multi-decade, multi-faceted campaign against protective orders.
In large measure, these efforts have focused on public access
questions,55 largely ignoring the impact of court confidentiality on
the litigation system. Up to this point, the fight has primarily
revolved around three axes: a First Amendment challenge, a
rulebased right of access argument, and transparency legislation.
One early line of attack against pretrial protective orders
was founded in the First Amendment.56 In the late 1970s,
protransparency advocates urged a First Amendment basis for
resisting protective orders: court orders restricting what litigants
do with information that they obtained in discovery allegedly
53. See, e.g., Joseph F. Anderson Jr., Hidden from the Public by Order of
the Court: The Case Against Government-Enforced Secrecy, 55 S.C. L. REV. 711,
715 (2004) (contending, that to a federal district judge, “courts too often
rubberstamp confidentiality orders presented to them, sometimes altogether ignoring
or merely giving lip service to the body of law and existing court rules that are
supposed to apply when the parties request that discovery documents be filed
under seal”); see also Pansy v. Borough of Stroudsburg, 23 F.3d 772, 785 (3d Cir.
1994) (observing the trend of some courts to enter protective orders as a matter
of course “without considering the propriety of such orders, or the countervailing
public interests which are sacrificed by the orders”).
54. See Model Protective Orders, U.S. DIST. COURT N. DIST. CAL.,
http://www.cand.uscourts.gov/model-protective-orders (last visited Nov. 19,
2014) (providing “court-approved model forms” for protective orders) (on file
with the Washington and Lee Law Review).
55. See, e.g., Zitrin, supra note 2, at 1567 (arguing in favor of transparency
on the grounds that “[o]nce the disputants go to court, the public nature of the
forum trumps the formerly private nature of the dispute”).
56. For a full analysis of the current relationship between the First
Amendment and pretrial protective orders, see generally Benham, supra note
46, at 1785–86.
violated the Constitution’s free-speech guarantee.57 Early cases
considering the argument suggested that confidentiality orders
did not violate the First Amendment.58
But in 1979, the D.C. Circuit entered the fray and set off a
firestorm by applying strict scrutiny to invalidate a protective
order.59 If this view prevailed, many protective orders would be
doomed. The Supreme Court, in Seattle Times v. Rhinehart,60
responded. Justice Lewis Powell, writing for a unanimous court,
rejected the D.C. Circuit’s strict-scrutiny approach and instead
held that, although the First Amendment did protect those who
disclose discovery information, the special context of civil
litigation implicated free-speech concerns to a lesser extent than
in other contexts.61 While courts and commentators continue to
debate the meaning of the case,62 multiple courts have held that
Seattle Times effectively removed the First Amendment from the
57. Cf., e.g., Michael Dore, Confidentiality Orders—The Proper Role of the
Courts in Providing Confidential Treatment for Information Disclosed Through
the Pre-Trial Discovery Process, 14 N. ENG. L. REV. 1, 10–14 (1978) (examining
the role of the First Amendment regarding protective orders).
58. See Int’l Prods. Corp. v. Koons, 325 F.2d 403, 407 (2d Cir. 1963) (“[W]e
entertain no doubt as to the constitutionality of a rule allowing a federal court to
forbid the publicizing, in advance of trial, of information obtained by one party
from another by use of the court’s process.”).
59. See In re Halkin, 598 F.2d 176, 183 (D.C. Cir. 1979) (“Even if the order
is relatively narrow, however, it restrains the petitioner from communicating
matters of public importance for an indefinite period of time. As such it
constitutes direct governmental action limiting speech and must be carefully
scrutinized in light of the first amendment.”).
476 U.S. 20 (1984).
61. See id. at 37 (“[W]here . . . a protective order is entered on a showing of
good cause as required by Rule 26(c), is limited to the context of pretrial civil
discovery, and does not restrict the dissemination of the information if gained
from other sources, it does not offend the First Amendment.”).
62. See, e.g., Benham, supra note 46, at 1804–14 (arguing that protective
orders implicate significant First Amendment concerns).
63. See Marcus, supra note 10, at 462–63 (“[A]ll are agreed that the broad
[First Amendment] prior restraint argument that captured attention in the
early 1980s no longer has force.”). Compare Cipollone v. Liggett Grp., Inc., 785
F.2d 1108, 1114 (3d Cir. 1986) (noting that “[d]espite the Supreme Court’s
apparent endorsement in the above passage of a least restrictive means
analysis, its holding requires only a good cause analysis”), with Anderson v.
Cryovac, Inc., 805 F.2d 1, 7 n.2 (1st Cir. 1986) (interpreting Seattle Times to
Soon after the Court decided Seattle Times, a few
commentators and courts cited the case to preclude discovery
sharing between similar litigants.64 That approach has been
roundly rejected because the case simply did not address the
propriety of sharing.65 Indeed, if anything, courts allowing
discovery sharing would impose a lesser speech restriction on
litigants than that approved by the Seattle Times court.66
On another front in the court-confidentiality fight,
transparency proponents have also contended that the Federal
Rules of Civil Procedure create a right of third-party access to
discovery materials.67 Under a previous version of Rule 5(d), all
discovery materials had to be filed with the court.68 The
reasoning went that the public had a common-law right to access
the unsealed contents of a court file.69 Because discovery was on
apply “heightened” scrutiny to the practice of issuing protective orders and
allowing some role for the First Amendment in the granting of particular
protective orders). But see Benham, supra note 46, at 1804–14 (“The Supreme
Court’s opinion in Seattle Times, recent Court statements, and an assessment of
the nature of both protective orders and the speech that they restrict make clear
that so-called intermediate scrutiny applies”).
64. See, e.g., Rich Arthurs, Defendants Fight Back on Data Sharing, LEGAL
TIMES, July 16, 1984, at 1 (describing Ford Motor Company’s attempt to use
trade secret protection to prevent plaintiffs’ attorneys from sharing crash test
data); Gary L. Wilson, Note, Seattle Times: What Effect on Discovery Sharing?,
1985 WIS. L. REV. 1055, 1056–57 (noting that the first court to apply Seattle
Times “relied on the rationales enunciated by the Supreme Court and granted a
protective order prohibiting the dissemination of discovered information to
parties involved in similar litigation”).
65. See, e.g., Marcus, supra note 10, at 495 (“[I]t is difficult to see how the
[Seattle Times] Court’s reasoning should prompt other courts to curtail access
for use in other litigation.”).
66. See Benham, supra note 46, at 1823 (explaining that courts utilize
sharing protective orders to allow litigants to share discovery with litigants in
67. See Seymour Moskowitz, Discovering Discovery: Non-Party Access to
Pretrial Information in the Federal Courts 1938–2006, 78 U. COLO. L. REV. 817,
865–75 (2007) (arguing that a presumption of access to discovery material was
“created by the combination of the Rule 5(d) filing requirement and the Rule 26
good cause provision”).
68. For a thorough history of changes to Rule 5(d) and the implications for
access arguments, see id. at 833–53.
69. See id. at 861–64 (observing that a general right to inspect and copy
both judicial and public records and documents existed).
file, or supposed to be, the public had a putative right to access
depositions, interrogatory responses, and other discovery
A series of rules amendments, culminating in a 2000
amendment that forbids parties from filing discovery unless in
connection with a proceeding or pursuant to a court order,71
effectively ended the Rule 5-based argument for a right of access
to unfiled discovery.72
Transparency, or “Sunshine,” legislation is yet another front
in the fight against protective orders. For more than a decade, the
Sunshine in Litigation Act has been filed, and re-filed, in both
branches of Congress but has never become law.73 Indeed, in May
2014, Senators Blumenthal and Graham reintroduced the
legislation as the Sunshine in Litigation Act of 2014 (SILA
The current act would require district courts to consider the
health and safety implications of both protective orders (even
agreed protective orders) and settlement agreements before
approving them.75 In particular, courts could not approve
70. See, e.g., Am. Tel. & Tel. Co. v. Grady, 594 F.2d 594, 596 (7th Cir. 1978)
(“As a general proposition, pretrial discovery must take place in the public
unless compelling reasons exist for denying the public access to the
FED. R. CIV. P. 5(d) (2000).
72. See, e.g., Marcus, supra note 50, at 333 (“[A]mendments to the Rules
that forbid the filing of most discovery in court . . . further support the proposal
that the notion of the Rules themselves commanding public access should be
laid to rest.”). But see generally Moskowitz, supra note 67 (indicating a
presumption of access to discovery materials).
73. See Mary Elizabeth Keaney, Note, Don’t Steal My Sunshine:
Deconstructing the Flawed Presumption of Privacy for Unfiled Documents
Exchanged During Discovery, 62 HASTINGS L.J. 795, 798 (2011) (noting that the
bill was originally introduced in 1993 by Senator Herbert Kohl, and it has been
reintroduced annually ever since).
S. 2364, 113th Cong. (2014).
See id. § 2 (as introduced by Senate, May 20, 2014)
Amends the federal judicial code to prohibit a court, in any civil
action in which the pleadings state facts relevant to protecting public
health or safety, from entering an order restricting the disclosure of
information obtained through discovery, approving a settlement
agreement that would restrict such disclosure, or restricting access to
court records, subject to exceptions, unless the court has first made
protective orders unless a “specific and substantial interest in
maintaining the confidentiality of the records” outweighed the
public interest in disclosure of “potential health or safety
Though primarily focused on giving the public access, SILA
2014 would also have implications for litigants seeking to share
information between related cases. If the public has access to
discovery materials so, presumably, would litigants in other
similar cases as members of the public. And the Act’s nexus to
public health and safety would often preclude protective orders in
the very class of cases that benefit most from discovery sharing—
products liability and tort actions involving products and injuries
of national proportions.77 As Congress considers SILA 2014, it
should consider the negative impact on court efficiency stemming
from protective orders along with the impact on public safety.
Additionally, some state legislatures have considered and
passed transparency and sunshine legislation, including at least
one statute that specifically addresses discovery sharing.78 Still,
the momentum for sunshine legislation in many states tapered
off in the mid-1990s.79
With the constitutional front in the fight against protective
orders largely dormant, the rule-based front effectively dead, and
independent findings of fact that: (1) the order would not restrict the
disclosure of information relevant to the protection of public health or
safety or (2) the public interest in the disclosure of past, present, or
potential health or safety hazards is outweighed by a specific and
substantial interest in maintaining the confidentiality of the
information and the requested protective order is no broader than
necessary to protect the confidentiality interest asserted.
77. See id. (prohibiting protective orders in cases “in which the pleadings
state facts that are relevant to the protection of public health or safety”).
78. See TEX. R. CIV. P. 76a (mandating a strict balancing test before courts
may enter protective orders restricting access to unfiled discovery that
implicates public health and safety); FLA. STAT. ANN. § 69.081 (West 2014)
(forbidding court orders restricting disclosure of information concerning public
hazards); VA. CODE ANN. § 8.01-420.01 (2014) (forbidding protective orders that
stifle discovery sharing).
79. But see MONT. CODE ANN. § 2-6-112 (2014) (forbidding, by statute
passed in 2005, certain court orders that have “the purpose or effect of
concealing a public hazard”).
the legislative front pending, the discovery-sharing discussion
often takes place in the context of trial court discretion.80 The
next subpart focuses on the relationship between protective-order
discretion and discovery sharing.
C. Sharing Discovery in Similar Cases
Litigants often seek to use discovery information obtained in
one case to prepare and develop another related case.81 The
concept known as “discovery sharing” flows from a basic,
uncontested proposition: Litigants may freely distribute discovery
information unless a valid court order forbids them from doing
so.82 And even if a court order limits how parties use discovery
information, judges have broad discretion to refuse dissemination
80. See, e.g., Ashley A. Kutz, Note, Rethinking The “Good Cause”
Requirement: A New Federal Approach to Granting Protective Orders Under
F.R.C.P. 26(c), 42 VAL. U. L. REV. 291, 293 (2007) (describing approaches to
determining good cause that range from “rubber-stamping” proposed protective
orders to “hostility toward confidentiality,” analyzing various circuits’
approaches to “good cause,” and proposing an improved protective-order process
81. See, e.g., Idar v. Cooper Tire & Rubber Co., No. C-10-217, 2011 WL
688871, at *3 (S.D. Tex. Feb. 17, 2011) (denying a protective order modification
while acknowledging that the order allows for “sharing of discovered
information with attorneys and other necessary persons representing parties
with present or future cases pending against Defendant that arise out of the
same or similar set of facts, transactions, or occurrences”); Wal-Mart Stores E.,
L.P. v. Endicott, 81 So. 3d 486, 490 (Fla. Dist. Ct. App. 2011) (noting
respondents’ request for a sharing provision to include collateral litigants);
Pincheira v. Allstate Ins. Co., 190 P.3d 322, 337 (N.M. 2008) (“The parties and
amici recognize that this case is really about discovery sharing, both with other
litigants and with the public at large.”); Cowan v. Gen. Motors Corp., No.
061330-MLB, 2007 WL 1796198, at *5 (D. Kan. June 19, 2007) (“The Court
recognizes the public policy benefit of potentially sharing documents with
attorneys involved in litigation of a similar product with similar issues.”).
82. See, e.g., Seattle Times Co. v. Rhinehart, 467 U.S. 20, 37 (1984)
(holding that where “a protective order is entered on a showing of good cause as
required by Rule 26(c), is limited to the context of pretrial civil discovery, and
does not restrict the dissemination of the information if gained from other
sources, it does not offend the First Amendment”); see also Moskowitz, supra
note 67, at 825 (“Absent a protective order, a plaintiff or defendant has the right
to disseminate information obtained during discovery so long as the purpose for
sharing is lawful.”).
to one group (e.g., the media) and allow dissemination to another
(e.g., litigants in similar, but separate, litigation).83
Plaintiffs’ attorneys typically contend that sharing
information between similar cases allows them to avoid
wastefully reinventing the wheel by repeatedly conducting
virtually identical discovery.84 Beyond these efficiency gains,
sharing proponents contend that exchanging information allows
isolated plaintiffs to prepare cases collaboratively, leveling the
playing field with large law firms and national collaboration on
the other side of the docket. Sharing also increases discovery
accountability by allowing parties to compare discovery responses
against responses in a similar case.85
Despite the apparent benefits, disputes about the propriety of
sharing have persisted for decades.86 Indeed, in the early 1990s,
one leading pro-confidentiality commentator had already noted
that the debate over so-called discovery sharing was
“retrograde.”87 Some courts’ continued reticence to allow it is
curious, particularly considering the overwhelming consensus
83. See, e.g., Raymond Handling Concepts Corp. v. Super. Ct., 45 Cal. Rptr.
2d 885, 890 (Cal. Ct. App. 1995) (noting that it is within trial courts’ discretion
to allow sharing with similar litigants while denying access to the general
84. See FRANCIS H. HARE, JR. ET AL., CONFIDENTIALITY ORDERS 69–70 (John
Wiley & Sons, Inc. 1988) (contending that discovery, as a result of nonsharing
protective orders, “must be repeated anew in every case,” with the trial courts
being forced to intervene in a repetition of the same discovery disputes); Ward v.
Ford Motor Co., 93 F.R.D. 579, 580 (D. Colo. 1982) (asserting that denying
sharing between litigants “would be tantamount to holding that each litigant
who wishes to ride a taxi to court must undertake the expense of inventing the
85. See, e.g., FRANCIS H. HARE, JR. ET AL., FULL DISCLOSURE: COMBATING
STONEWALLING AND OTHER DISCOVERY ABUSES 166 (ATLA Press 1995) (arguing
that “information sharing also provides plaintiffs the opportunity to verify the
completeness and accuracy of a defendant manufacturer’s response to discovery
request”); cf. Marcus, supra note 10, at 495–96 (“In addition, in some cases
access may prevent efforts to mislead the court in the second case.”).
86. Compare Campbell, supra note 2, at 824 (opposing information sharing
on the ground that it creates waste and threatens proprietary interests of
parties), with Marcus, supra note 10, at 495–96 (observing that discovery
sharing benefits the litigation system), and Rhiana Sharp, Comment, Let in a
Little Sunshine: Limiting Confidential Settlements in Missouri, 69 MO. L. REV.
215, 229 (2004) (“Discovery sharing should be encouraged.”).
Marcus, supra note 10, at 495.
is discoverable in their own case.246 The approach is rooted in
comity concerns and concerns about forum shopping.247
Indeed, several courts have resisted sharing orders on the
ground that sharing provisions have the potential to thwart
the discovery limitations of the jurisdictions in which
litigants receive shared information.248 According to this line
of reasoning, crafty litigants stuck in a jurisdiction with
limited discovery could find a similar case with a sharing
order in a jurisdiction with more liberal discovery standards.
By doing so, the litigant might obtain information to which he
or she would not otherwise be entitled. Accordingly, some
courts deny sharing requests in favor of litigant-by-litigant
This process and the reasoning underlying it are flawed for
several reasons. Even if an upfront sharing provision ignores
discovery limitations in a collateral case, it does not harm the
246. See, e.g., Marcus, supra note 2, at 43 (arguing that third-party litigant
seeking to modify protective order should establish that “he would have the
right to obtain” discovery in collateral action); 8A WRIGHT ET AL., supra note 16,
§ 2044.1 (“If the limitation on discovery in the collateral litigation would be
substantially subverted by allowing access to discovery material under a
protective order, the court should be inclined to deny modification.”); cf. Foltz v.
State Farm Mut. Auto. Ins. Co., 331 F.3d 1122, 1132 (9th Cir. 2003) (“[T]he
collateral litigant must demonstrate the relevance of the protected discovery to
the collateral proceedings and its general discoverability therein. Requiring a
showing of relevance prevents collateral litigants from gaining access to
discovery materials merely to subvert limitations on discovery in another
247. See, e.g., Foltz, 331 F.3d at 1133 (“Allowing the parties to the collateral
litigation to raise specific relevance and privilege objections to the production of
any otherwise properly protected materials in the collateral courts further
serves to prevent the subversion of limitations on discovery in the collateral
248. See, e.g., Bertetto v. Eon Labs, Inc., No. 06-1136 JCH/ACT, 2008 WL
2522571 (D.N.M. May 29, 2008) (stating that sharing provisions would
“potentially usurp . . . a collateral court’s role in managing discovery in a
collateral case”); see also Cordis Corp. v. O’Shea, 988 So. 2d 1163, 1167 (Fla.
Dist. Ct. App. 2008) (“Allowing this degree of sharing of confidential information
may provide a mechanism for attorneys in states with narrower discovery laws
to evade their state law discovery limitations . . . .”).
249. See, e.g., Cordis Corp., 988 So. 2d at 1167 (noting that “not every
federal court decision has approved sharing confidential information even with
collateral litigants or counsel”).
producing party in a way recognized by Rule 26(c). When
previously produced information is shared, the producing party
has presumably already expended resources to complete the
discovery in the sharing jurisdiction. The collateral party
receiving the shared information is responsible to pay to copy the
documents.250 Moreover, the party benefiting from sharing will
almost always be required to agree to be bound by the terms of
the protective order (e.g., use the documents only to prepare and
try the collateral case).251 If anything, the producing party has
saved expense and time in the collateral case.252 The system has
undoubtedly benefited, particularly if the party receiving the
shared information is able to exercise her discretion better in
framing discovery requests in the collateral case or avoid some
Perhaps the producing party could complain that the
collateral party received undue access to information that helped
prove her case. But discovery limitations are meant only to
balance the search for truth against the cost of that search, not to
serve as a shield against liability.253 Moreover, the scope of
discovery is necessarily broader than the standard for
admissibility.254 If shared discovery yields relevant, admissible
250. See FED. R. CIV. P. 26 advisory committee’s note (“The conditions may
also include payment by the requesting party of part or all of the reasonable
costs of obtaining information from sources that are not reasonably
251. See Raymond Handling Concepts Corp. v. Super. Ct., 45 Cal. Rptr. 2d
885, 886 (Cal. Ct. App. 1995) (describing a protective order with a typical
sharing provision, tailored to ensure compliance by counsel receiving the shared
252. See Ward v. Ford Motor Co., 93 F.R.D. 579, 580 (D. Colo. 1982)
(observing that sharing protective orders save defendant expense in time that
would otherwise be spent on duplicative discovery in similar cases).
253. Cf. Hickman v. Taylor, 329 U.S. 495, 507–08 (1947) (delineating
circumstances in discovery which could cause “the rights of individual litigants
[to be] drained of vitality and the lawsuit [to become] more of a battle of
deception than a search for truth”).
254. See FED. R. CIV. P. 26(b)(1) (allowing “[p]arties [to] obtain discovery
regarding any nonprivileged matter that is relevant to any party’s claim or
defense” even inadmissible information).
evidence, the receiving jurisdiction’s trial process has been
enhanced, not thwarted.255
Indeed, it is difficult to imagine what complaint the court in
the receiving jurisdiction might have. If anything, the court in the
receiving jurisdiction has had its interests furthered by the
sharing protective order, particularly if the sharing jurisdiction
has allowed broad discovery.256 With more proof available,
attorneys may be less likely to engage in discovery disputes in the
receiving jurisdiction, saving resources for that court. And more
information, rather than less, is almost always beneficial in
seeking truth.257 When that information comes to the collateral
court essentially free of cost via a previously litigated case, it is
hard to imagine the downside.
One legitimate concern, albeit unlikely to occur often, is that
attorneys engaged in national-scale claims might seek out a
jurisdiction with unusually liberal discovery standards and create
a discovery mill for other claims around the country.258
Presumably, the attorneys could file an action, obtain favorable
discovery rulings, obtain a sharing protective order, and then use
the jurisdiction to feed cases around the country, notwithstanding
the merits of the collateral sharing case.
At the outset, there are almost no identified instances of this
hypothetical occurring. In the event that a court had evidence of
such a scheme, it would have ample discretion to deny a sharing
255. See Garcia v. Peeples, 734 S.W.2d 343, 347 (Tex. 1987) (“Shared
discovery is an effective means to insure full and fair disclosure. Parties subject
to a number of suits concerning the same subject matter are forced to be
consistent in their responses by the knowledge that their opponents can
compare those responses.”).
256. Cf. Patterson v. Ford Motor Co., 85 F.R.D. 152, 154 (W.D. Tex. 1980)
(“The availability of the discovery information may reduce time and money
which must be expended in similar proceedings, and may allow for effective,
speedy, and efficient representation.”).
257. See Garcia, 734 S.W.2d at 347 (“[T]he ultimate purpose of discovery is
to seek the truth, so that disputes may be decided by what the facts reveal, not
by what facts are concealed.”).
258. See Cordis Corp. v. O’Shea, 988 So. 2d 1163, 1167 (Fla. Dist. Ct. App.
2008) (warning that permitting parties to share confidential information “may
provide a mechanism for attorneys in states with narrower discovery laws to
evade their state law discovery limitations by obtaining confidential
information” from a more liberal jurisdiction).
provision in the first place or modify an existing protective order
to shut the discovery mill down.259 Utilizing a case primarily as a
vehicle for obtaining information to overrun other jurisdiction’s
discovery limitations—rather than to work toward a resolution on
the merits—would be inappropriate.260 But the mere specter of
such abuse does not constitute reason to abandon discovery
sharing in favor of ad hoc modification, particularly when courts
have ample tools to deal with any instances that do arise.261
Nevertheless, the discoverability requirement is a predicate
to sharing in many jurisdictions.262 As a result, parties forced to
seek modification of protective orders may, in essence, litigate
disputes about discoverability substantially similar to those they
would face in their own cases. Upfront sharing provisions could
avoid the problem altogether.
Beyond the direct efficiencies generated by sharing
provisions, the provisions also provide some indirect, systemic
benefits. Indeed, up-front sharing provisions encourage party
cooperation and efficiency in the long run by avoiding reliance
problems and properly setting party expectations at the inception
of protective orders.263 Later protective-order modifications often
fall short in this regard.264
259. See, e.g., Patterson v. Ford Motor Co., 85 F.R.D. 152, 154 (W.D. Tex.
1980) (“Unless it can be shown that the discovering party is exploiting the
instant litigation solely to assist litigation in a foreign forum, federal courts
allow full use of the information in other forums.”).
260. See id. (highlighting the primary aim of the Federal Rules of Civil
Procedure: to “secure the just, speedy, and inexpensive determination of every
261. See FED. R. CIV. P. 37 (permitting courts to sanction parties for failing
to comply with court orders); FED. R. CIV. P. 26(g)(3) (providing that a court may
impose appropriate sanctions on a party or party’s attorney for violating the
certification standards for discovery requests and disclosures).
262. See, e.g., Long v. TRW Vehicle Safety Sys., Inc., No.
CV-09-2209-PHXDGC, 2010 WL 1740831, at *1 (D. Ariz. Apr. 29, 2010) (“[T]he collateral litigant
must demonstrate the relevance of the protected discovery to the collateral
proceedings and its general discoverability therein.”).
263. See, e.g., Diversified Grp., Inc. v. Daugerdas, 217 F.R.D. 152, 158
(S.D.N.Y. 2003) (noting one court’s “restrictive attitude toward modification of
protective orders” and arguing that “[i]f protective orders were easily
modified, . . . parties would be less forthcoming in giving testimony and less
willing to settle their disputes”).
264. See, e.g., id. (“Where there has been reasonable reliance by a party on a
Courts considering protective-order modification requests
regularly struggle with questions about reliance.265 Courts often
hesitate to grant modifications to allow sharing after the parties
have produced information to each other in reliance on a
confidentiality order.266 The basis for this hesitation is that the
parties produced the information with the expectation that it
would be used only for the case at hand.267
By defeating party expectations about confidentiality, so the
logic goes, litigants will be less likely to produce information
under protective orders in the future, causing inefficiency and
other problems for the discovery system as a whole.268 As
discussed in Part IV.B, this reasoning may overstate the impact
of defeated protective order expectations.
Whatever the true extent and impact of the reliance problem,
sharing protective orders avoid questions about reliance
altogether in most cases.269 When a court enters a protective
order with a well-crafted sharing provision, all parties to the
order know, to a large extent, who may receive information under
protective order, a district court should not modify the order ‘absent a showing
of improvidence in the grant of [the] order or some extraordinary circumstance
or compelling need.’”).
265. See Olympic Ref. Co. v. Carter, 332 F.2d 260, 264 (9th Cir. 1964)
(noting acts committed by the defendants “in reliance upon the entry and
continuing effectiveness of the protective orders”), cert. denied, 379 U.S. 900
266. See Martindell v. Int’l Tel. & Tel. Corp., 594 F.2d 291, 293 (2d Cir.
1979) (rejecting the Government’s modification request due to the opposing
party’s reliance on the prior order).
267. See Campbell, supra note 2, at 772–75 (explaining that “[t]he number
and variety of potential adversaries may make the ramifications of unfettered
dissemination of discovery information overwhelming”); Marcus, supra note 2,
at 43–44 (arguing that “a court should ordinarily deny nonparty access if all the
parties to litigation number one oppose it, even though it may increase the
268. See Campbell, supra note 2, at 772–75 (addressing potential risks to
litigants resulting from “[t]he uncontrolled dissemination of discovery
information among an ill-defined and untraceable amalgam of adversaries” and
why this would lead to lessened efficiency); Marcus, supra note 2, at 43–44
(noting that parties who foresee repeated claims may “fight discovery more
269. See Timmins, supra note 92, at 1520 (noting that “protective orders
that allow information sharing can substantially reduce the burdens that the
discovery process imposes on the plaintiff”).
that order. The increased certainty flowing from settled party
expectations about litigant-to-litigant sharing may even, by itself,
increase party willingness to provide forthright and complete
In sum, in many cases, particularly products-liability and
mass tort cases, later requests from collateral litigants to modify
protective orders are perfectly foreseeable and avoidable.271
Courts should set reasonable party expectations by entering
sharing protective orders that account for this reality at the
For some courts, the practice of sharing privileged
information across jurisdictions presents special difficulties.
Privileges are often based on policy judgments meant to influence
real-world behavior.273 They represent particular jurisdictions’
views on how to best maintain and further society, not simply run
a court system.274 These policy views manifest themselves in
privileges by limiting litigant and court access to information.
In contrast, Rule 26(c)(1)(g) is aimed at limiting the audience
for materials that have already been, or will be, found
discoverable.275 The focus of the protective order rules across
jurisdictions (largely based on Rule 26(c)) is unquestionably
270. See Garcia v. Peeples, 734 S.W.2d 343, 347–48 (Tex. 1987) (explaining
that the goals of discovery are “often frustrated by the adversarial approach to
discovery,” and that “[s]hared discovery is an effective means to insure full and
271. See Campbell, supra note 2, at 827 (“Particularly in products liability
cases, plaintiffs often routinely move to modify protective orders at the close of
litigation . . . .”).
272. See id. at 831 (criticizing umbrella protective orders which do not
account for this possibility).
273. See 2 CHRISTOPHER B. MUELLER & LAIRD C. KIRKPATRICK, FEDERAL
EVIDENCE § 5:2 (4th ed. 2013) (“Privilege rules occupy a unique place because
they implement policies that are very different from those underlying other
rules, and many privileges are broader in effect than other rules.”).
274. See id. (noting that privileges “are not designed to enhance the
reliability of factfinding, and they exclude evidence that may be probative
because other values are more important than finding truth in litigation”).
275. See FED. R. CIV. P. 26(c)(1)(G) (allowing courts to issue a protective
order “requiring that a trade secret or other confidential research, development,
or commercial information not be revealed or be revealed only in a specified
important but is simultaneously more value-neutral and more
homogenous on value questions than the privilege rules.276
For instance, the spousal privilege reflects the judgment that
marriage is a good that the government should promote.277 The
clerical privilege recognizes that candor between priest and
parishioner, and the spiritual enlightenment that might follow, is
good for society.278
As one might expect, the disagreements about the value
judgments underlying privileges have produced substantial
disparity in privilege law. Some jurisdictions recognize the
physician–patient privilege while others do not.279 Even when
lawmakers in different jurisdictions agree about the existence of
a privilege, the contours of its applicability may vary
dramatically.280 Adding to the complexity in the area, privileges
are largely the province of state law in civil cases, even in federal
courts.281 The fifty states have come to widely varying conclusions
on privilege questions, an unsurprising result that reflects the
political and moral diversity of the country.
276. See, e.g., Sanford Levinson, Testimonial Privileges and the Preferences
of Friendship, 1984 DUKE L.J. 631, 662 (stating that privileges “lie at the center
of the contemporary debate about the foundations of liberal society” and that
“[t]o privilege certain relationships is to declare certain values”).
See Hawkins v. United States, 358 U.S. 74, 77 (1958)
The basic reason the law has refused to pit wife against husband or
husband against wife in a trial where life or liberty is at stake was a
belief that such a policy was necessary to foster family peace, not only
for the benefit of husband, wife and children, but for the benefit of the
public as well.
278. See Trammel v. United States, 445 U.S. 40, 51 (1980) (“The
priestpenitent privilege recognizes the human need to disclose to a spiritual
counselor, in total and absolute confidence, what are believed to be flawed acts
or thoughts and to receive priestly consolation and guidance in return.”).
279. Compare CAL. EVID. CODE § 994 (West 2008) (describing physician–
patient privilege), with United States v. Moore, 970 F.2d 48, 50 (5th Cir. 1992)
(“This Court has . . . concluded that there is no doctor-patient privilege under
280. Compare TEX. R. EVID. 509 (applying the physician–patient privilege in
civil, not criminal, cases), with CAL. EVID. CODE § 300, 994 (applying the
physician–patient privilege in criminal and civil cases).
281. See FED. R. EVID. 501 (“But in a civil case, state law governs privilege
regarding a claim or defense for which state law supplies the rule of decision.”).
But these value judgments are not implicated by
protectiveorder procedures. If a privilege applies and prevents disclosure in
discovery, an audience-limiting protective order simply is not
necessary. Indeed, if a privilege keeps information out of
discovery, there is simply nothing to protect.
On the other hand, if one state’s privilege law in a particular
area (or lack thereof) allows disclosure of information that would
be privileged in another state, and no court enters a protective
order, the first state’s privilege law may thwart the second’s.282
The same is true if a court in a state where information is not
privileged allows discovery sharing with cases in states where
privilege would limit disclosure. The information still would not
reach the general public (because the collateral party presumably
received the information by agreeing to the protective-order’s
restrictions against further disclosure); but it would be available
for litigation purposes, contrary to the receiving jurisdiction’s
One response to this might be simply that the states often
thwart one another’s policy judgments, an undeniable reality of a
multi-sovereign republic. The disparity in privilege law, largely
recognized as appropriately within the discretion of state
lawmakers or courts, might be more the culprit here than the
sharing order. Indeed, the court in the sharing jurisdiction was
not obligated to enter a protective order at all.284 And if it had
refused to grant any confidentiality, its previous privilege ruling
would have, in itself, thwarted the policy of the receiving
jurisdiction by allowing would-be privileged information to flow
In this same way, Colorado’s policy legalizing marijuana
thwarts the criminal justice policy of its neighbor states when
282. See 2 MUELLER & KIRKPATRICK, supra note 273, § 5.8 (delineating which
state’s privilege laws apply).
283. The party holding the privilege may, of course, still contest the
admissibility of the information at trial. See, e.g., Marcus, supra note 10, at 496
(“[S]haring is consistent with a settled doctrine exempting exchange of
information between litigants with a common litigation opponent from the risk
that the exchange will be deemed a waiver of privileges.”).
284. See, e.g., Seattle Times Co. v. Rhinehart, 467 U.S. 20, 36 (1984) (“Rule
26(c) confers broad discretion on the trial court to decide when a protective order
is appropriate and what degree of protection is required.”).
Colorado-grown pot flows across the border.285 Rather than
serving as a reason to limit discovery sharing, tension among
states’ privilege law may simply be an inevitable by-product of
states’ recognized ability to make their own policy choices.
Another version of the privilege problem comes up when
sharing courts order parties to produce information pursuant to a
qualified privilege.286 The trade-secret privilege provides an apt
example. Many states qualify the trade secret privilege to allow
courts to compel production of otherwise privileged information
where failing to do so would “conceal fraud” or otherwise create
“an injustice.”287 The burden is often placed on the party seeking
the privileged information to establish that injustice or
concealment of a fraud would occur absent disclosure in the
Some argue that when courts allow sharing in trade-secret
cases without requiring such a showing, the protective-order
ruling might actually conflict with the sharing jurisdiction’s own
privilege law.289 Many states’ privileges not only allow the party
holding the privilege to refuse to provide the information absent a
showing of injustice or fraud, but also to prevent others from
doing so.290 A sharing protective order in such a jurisdiction,
285. See, e.g., Jack Healy, After 5 Months of Sales, Colorado Sees the
Downside of a Legal High, N.Y. TIMES (May 31, 2014), http://nyti.ms/1kjtwf8
(last visited Nov. 19, 2014) (highlighting that some law enforcement officials in
neighboring states that have seen an increase in marijuana-related offenses) (on
file with the Washington and Lee Law Review).
286. See Wal-Mart Stores E., L.P. v. Endicott, 81 So. 3d 486, 490 (Fla. Dist.
Ct. App. 2011) (addressing the tension between sharing protective orders and
state trade-secret privilege).
288. See, e.g., id. (stating that “sharing provisions, like the underlying one,
which allow the dissemination of trade secrets without considering these factors
codified” under Florida law “are per se unlawful”).
289. See, e.g., FLA. STAT. ANN. § 90.506 (West 1995) (attempting to
circumvent the conflict between disclosure and privilege by directing courts to
“take protective measures that the interests of the holder of the
privilege . . . require” when directing disclosure of such holder’s trade secret).
290. See, e.g., FLA. STAT. ANN. § 90.506 (“A person has a privilege to refuse to
disclose, and to prevent other persons from disclosing, a trade secret owned by
that person if the allowance of the privilege will not conceal fraud or otherwise
work injustice.”); see also, e.g., Seattle Times Co. v. Rhinehart, 467 U.S. 20, 37
(1984) (holding that a protective order “entered on a showing of good cause as
according to some, could potentially undermine the privilege in
all future cases within the sharing provision after the first
But in the trade-secret-privilege context, these concerns
conflate two separate issues and overstate the concern. First, the
question of whether material is discoverable or not because of a
privilege is distinct from concerns about limiting the proper
audience for the material through protective orders.292 If a court
orders a party to produce otherwise privileged information
because failing to do so would be “an injustice,” and does not
enter a protective order, the information could be disseminated to
the general public.
Of course, the inquiries inform one another. A court should
consider the efficacy of protective measures when deciding to
order production of trade secrets or other privileged
information.293 Whether to enter such an order, however, is a
Rule 26(c) question, subject to the trial court’s broad discretion,
not a privilege question. This distinction means that Rule 26(c)’s
good cause standard, not privilege law, governs the size of the
audience for trade-secret information a court has deemed
Moreover, protective orders that account for similarity
between the sharing case and receiving case adequately account
for trade secret holders’ interests. Well-crafted sharing orders
allow sharing only with factually similar litigation.294 Thus, the
required by Rule 26(c), . . . limited to the context of pretrial civil discovery, [that]
does not restrict the dissemination of the information if gained from other
sources” does not violate the First Amendment); Moskowitz, supra note 67, 824–
25 (discussing the role of good cause in granting protective orders).
291. See Wal-Mart Stores, 81 So. 3d at 490 (“Sharing provisions, like the
underlying one, which allow the dissemination of trade secrets without
considering [the operation of the privilege] are per se unlawful.”).
292. Cf., e.g, Chapa v. Garcia, 848 S.W.2d 667, 673 n.8 (Tex. 1992) (Doggett,
J., concurring) (“While discoverability by the parties is often confused with
disclosability to the public, discoverability and disclosability issues must be
293. See In re Cont’l Gen. Tire, Inc., 979 S.W.2d 609, 613 (Tex. 1998)
(holding that courts should enter appropriate protective order after determining
whether to order production of trade secret information).
294. See, e.g., Raymond Handling Concepts Corp. v. Super. Ct., 45 Cal. Rptr.
2d 885, 886 (Cal Ct. App. 1995) (providing an example of a protective order that
shared information is likely necessary and relevant in the
receiving case, satisfying the “injustice” prong of the trade-secret
privilege.295 If the discovery is relevant and necessary in one case
involving a defective widget that injured a person it likely has a
sufficient connection to another case involving the same widget
that injured another person.
In essence, if the court compels disclosure of trade-secret
information and simultaneously grants a protective order that
allows sharing with similar cases, it has implicitly ruled that the
information is necessary for the proper preparation of all cases
within the sharing class.
To the extent the court has concerns about the relevance and
necessity in cases receiving the information, it can address those
concerns by narrowing the sharing class, not eliminating it
entirely. An approach that allows appropriate sharing of
tradesecret information, pursuant to a thoughtful and well-crafted
sharing order, is consistent with the qualified trade-secret
Second, courts have long recognized that the trade-secret
privilege is qualified rather than absolute.296 Indeed, prominent
commentators have recognized that protection for trade secrets is
not a “true” privilege because trade secrets are “generally
discoverable where they are relevant to the dispute.”297 In many
jurisdictions, the showings necessary to trigger compelled
disclosure of trade secrets because of “injustice” boil down to
something more than, but akin to, relevance.298 Courts take this
allowed sharing of information “in other pending similar litigation”).
295. See FLA. STAT. ANN. § 90.506 (stating that information may only fall
within the trade-secret provision “if the allowance of the privilege will
not . . . otherwise work an injustice”).
296. See, e.g., Bleacher v. Bristol-Myers Co., 163 A.2d 526, 529 (Del. Super.
Ct. 1960) (“Before a court will order a manufacturer to divulge information
relating to a secret process or formula of his product, it must be clearly shown
that the information required is relevant to the issue, not otherwise available
and necessary in the proof of plaintiff’s case.”).
297. 2 MUELLER & KIRKPATRICK, supra note 274, § 5:49.
298. See, e.g., Bleacher, 163 A.2d at 529 (finding the plaintiff entitled to the
trade secret information necessary to prove her case); see also 2 MUELLER &
KIRKPATRICK, supra note 274, § 5:49 (“The term ‘privilege’ appears in this
setting, but there is no ‘true’ privilege simply because trade secrets and
competitive information are generally discoverable where they are relevant to
lenient approach because the trade-secret privilege works in
tandem with protective orders to keep information from
The purpose of the privilege is to preserve the value of
tradesecret information by forbidding its disclosure to the privilege
holder’s economic foes.299 Almost all sharing orders forbid
dissemination to the disclosing party’s competitors. By doing so,
the orders protect the value of the disclosing party’s secrets. It is
well established that compelling the disclosure of trade secrets in
litigation subject to a protective order that prevents competitors
from accessing the information is appropriate.300
B. Modification Standards
To the extent courts use nonsharing protective orders,
collateral litigants should be able to obtain easy access to
discovery for use in related cases. Most courts to consider the
issue take this liberal approach, championed by Wilk v. American
Medical Association.301 Wilk’s presumption in favor of modifying
the dispute.”). But see In re Cont’l Gen. Tire, Inc., 979 S.W.2d at 613–14
(rejecting the argument that mere relevance is enough to overcome qualified
trade secret privilege); cf. 26 WRIGHT ET AL., FEDERAL PRACTICE AND PROCEDURE
§ 5651 (1st ed. 2010) (stating that the concept of “injustice” in the qualified
trade-secret context “expanded in the caselaw to any situation in which the facts
cannot be ascertained without disclosure of the trade secrets” (internal
quotation marks and citation omitted)).
299. See, e.g., In re Bass, 113 S.W.3d 735, 739 (Tex. 2003) (“[A] trade secret
is any formula, pattern, device or compilation of information which is used in
one’s business and presents an opportunity to obtain an advantage over
competitors who do not know or use it.” (internal quotation marks and citation
300. See, e.g., Fed. Open Mkt. Comm. of Fed. Reserve Sys. v. Merrill, 443
U.S. 340, 362 n.24 (1979) (“Actually, orders forbidding any disclosure of trade
secrets or confidential commercial information are rare. More commonly, the
trial court will enter a protective order restricting disclosure to counsel. . . .”).
301. 635 F.2d 1295, 1299 (7th Cir. 1980) (“[W]here an appropriate
modification of a protective order can place private litigants in a position they
would otherwise reach only after repetition of another’s discovery, such
modification can be denied only where it would tangibly prejudice substantial
rights of the party opposing modification.”), superseded by rule as stated in Bond
v. Utreras, 585 F.3d 1061 (7th Cir. 2009); see also, e.g., Foltz v. State Farm Mut.
Auto. Ins. Co., 331 F.3d 1122, 1131–32 (9th Cir. 2003) (citing Wilk with approval
protective orders to aid collateral litigation best supports Rule 1
interests, as highlighted at the Duke Conference and in the
proposed civil rules amendments.302
Some courts, however, disagree and continue to impose an
unduly high barrier to modifying protective orders.303 The
primary justification for doing so is the supposed harm to parties
that have produced information in reliance on the order.304 But
reality belies this reasoning. First, many protective orders are
stipulated umbrella orders, rubberstamped by a court.305 By their
very terms, these orders allow any party to the case to challenge
the “confidentiality” of a document.306 If a court sustains the
challenge, the document is immediately stripped of protection. It
is hard to envision how a party might view such an order as a
guarantee of confidentiality and thereby rely on it.307 Any party
producing putatively confidential documents under an umbrella
and requiring something less than Martindell’s extraordinary circumstances
test); Pansy v. Borough of Stroudsburg, 23 F.3d 772, 790 (3d Cir. 1994)
(rejecting extraordinary circumstances test as too strict); United Nuclear Corp.
v. Cranford Ins. Co., 905 F.2d 1424, 1428 (10th Cir. 1990) (adopting the
standard in Wilk); Comes v. Microsoft Corp., 775 N.W.2d 302, 309–10 (Iowa
2009) (adopting the less stringent test from Wilk).
302. See COMMITTEE REPORT, supra note 6, at 77 (amending Rule 1 to
emphasize that the parties, as well as the courts, have the responsibility to
“secure the just, speedy, and inexpensive determination of every action”).
303. See Martindell v. Int’l Tel. & Tel. Corp., 594 F.2d 291, 296 (2d Cir.
1979) (requiring “a showing of improvidence in the grant of a Rule 26(c)
protective order or some extraordinary circumstance or compelling need” to
disclose information under the umbrella of a protective order).
304. See id. (rejecting a government attempt to modify a protective order
where parties relied on the order in providing testimony that would otherwise
be privileged under the 5th Amendment).
305. See Howard M. Erichson, Court-Ordered Confidentiality in Discovery,
81 CHI.-KENT L. REV. 357, 357 (2006) (“Disturbingly, some courts routinely sign
orders which contain confidentiality clauses without considering the propriety of
such orders, or the countervailing public interests which are sacrificed by the
orders.” (citing Pansy v. Borough of Stroudsburg, 23 F.3d 772, 785 (3d. Cir.
306. See Marcus, supra note 10, at 500–02 (describing, with approval, the
operation of umbrella protective orders).
307. See id. at 502 (“Admittedly, the court may find the fact that an
umbrella order has been used bears on reasonable reliance issues if modification
of the order is sought. . . .”).
order is taking a leap of faith, not relying on a promise of future
Second, it is well known that courts frequently modify
protective orders to allow third parties to use discovery
information in similar collateral litigation, particularly in
products-liability litigation.308 Indeed, many protective orders
specifically indicate that the court retains jurisdiction to modify
the order even after the primary litigation has ended. To the
extent any party relies on an order expressly indicating it is
subject to modification in a litigation system that frequently
modifies orders for litigation-related use, such reliance is
Even if a party relies on a protective order for confidentiality,
any harm flowing from a modification to allow information to be
used in related litigation is minimal or nonexistent in most cases.
Only two sources of harm typically flow from sharing-related
modifications. Neither justifies an onerous modification standard.
First, a party that produced inculpatory information in the
sharing case may be held accountable in a receiving case for the
real-world misconduct the shared information reveals. While this
is undoubtedly undesirable for the producing party, courts should
not resist modifying protective orders to aid parties in escaping
liability in subsequent cases.309 It is true that litigants sometimes
escape liability as a by-product of the rules’ goals (e.g., when a
rule precludes disclosure of a damning fact for the sake of
minimizing discovery’s burdens or protecting another process),
but minimizing liability for misconduct is not an appropriate use
of the discovery rules.310
Many courts have lauded discovery sharing’s capability to
hold defendants accountable for varying discovery responses
308. See id. at 498 (acknowledging courts’ willingness to modify orders
makes reliance less reasonable in many cases).
309. Cf. Dore, supra note 1, at 363 (arguing that protective orders should not
be used to shield against liability).
310. See also United States v. Procter & Gamble Co., 356 U.S. 677, 682
(1958) (discussing the great need for grand jury transcripts to be kept
confidential); cf. Dore, supra note 1, at 363 (arguing that protective orders
should not be used as shield against liability).
across different cases.311 Better holding defendants accountable
for the real-world conduct that led to litigation seems to be an
even more fundamental benefit of shared discovery. To the extent
parties produce information in one case in reliance on a protective
order to shield them from an inculpatory disclosure in a future
case, their reliance is not detrimental in a sense that Rule 26
A second harm Rule 26 might recognize is an increased risk
of public disclosure, potentially undermining the competitive
value of information.313 When a protective order is modified, the
audience of litigants and attorneys widens. Each additional
person who receives the information makes it slightly more likely
that someone violates the order’s terms and disseminates the
information to a broad, even public, audience.314 The potential for
mass dissemination may arguably be more acute in the Internet
age, when one rogue recipient could disseminate an entire
discovery production to a global audience.315
Nevertheless, experience has shown that the risk of such
misconduct is minimal. During the more than forty years of the
modern protective-order regime (including almost twenty years in
311. See In re Abbott Labs. Sec. Litig., No. 92 C 3869, 1993 WL 616693, at
*4 (N.D. Ill. Nov. 15, 1993) (stating that discovery sharing “allows plaintiffs’
counsel to verify that they, in their respective cases, have in fact received all of
the documentation they are entitled to”).
312. But see Martindell v. Int’l Tel. & Tel. Corp., 594 F.2d 291, 295–96 (2d
Cir. 1979) (finding that the district court erred in allowing the Government to
obtain protected testimony because “a witness should be entitled to rely upon
the enforceability of a protective order against any third parties” otherwise such
witnesses would be “inhibited from giving essential testimony in civil
313. See Campbell, supra note 2, at 824 (discussing the potential for “serious
problems of policing and enforcement” of protective orders and noting that
“[o]nce the information has been transferred beyond the limits of the jurisdiction
where it was originally discovered, the court’s contempt power will no longer
reach those in its possession”).
314. See id. (“[T]he likelihood of violation, inadvertent or otherwise, will
increase in direct proportion to the number of disclosures.”).
315. See Marder, supra note 26, at 318 (observing that with the existence of
the Internet “the reach of ‘public’ has expanded, and the consequences of any
mistaken disclosure are far greater than when the ‘public’ meant an individual
who went down to the courthouse . . . and then returned the document” after
the Internet age), notable protective order violations are
relatively rare.316 The relatively low number of serious violations
likely stems from the serious consequences imposed on attorneys
who violate orders. Contempt is one possible sanction.317 A less
recognized informal punishment, but one that may serve as a
serious deterrent, is the likely use of an attorney’s
protectiveorder violation as evidence in other courts to prevent that same
attorney, or expert, from obtaining access to discovery in future
cases.318 This informal sanction could undermine a
protectiveorder violator’s livelihood.
Moreover, harm flowing from reliance on an order modified
to allow yet-to-occur future violations is speculative.319 When
courts assess the detriments of a modification, they should focus
on actual or likely harm, not low-level possibilities.320
Protectiveorder violations are rare, but unauthorized disclosure to a
competitor of a producing party as the result of a protective-order
modification is almost unheard of.
316. Some violations have occurred over the years, but courts have displayed
robust willingness to enforce their orders. Relative to the tens of thousands of
protective orders issued during the Rule 26(c) era, the number of violations
remains small. But see Smith & Fuller, P.A. v. Cooper Tire & Rubber Co., 685
F.3d 486, 490 (5th Cir. 2012) (upholding sanctions for inadvertent violation
when attorneys had unintentionally disseminated materials via a CD provided
to a plaintiffs’ conference audience); McDonald v. Cooper Tire & Rubber Co., 186
F. App’x 930, 932 (11th Cir. 2006) (finding violation of protective order); Nevil v.
Ford Motor Co., No. CV 294-015, 1999 WL 1338625, at *4 (S.D. Ga. Dec. 23,
1999) (finding expert in contempt when expert disclosed information about
protected documents in collateral litigation).
317. See, e.g., Marcus, supra note 50, at 348 (noting that “contempt of court”
is a potential punishment for violating a protective order).
318. Defendants are not shy about using past protective-order violations
against the violator to urge courts not to grant access to information. See
Defendant Cooper Tire & Rubber Co.’s Memorandum of Law in Opposition to
Plaintiffs’ Motion to Compel & in Support of Cooper’s Motion for a Protective
Order at 45, Richards v. Ford Motor Co., No. 310958/2011, 2013 WL 6096311
(N.Y. Sup. Ct. Dec. 9, 2011) (emphasizing that the “[p]laintiffs’ retained tire
expert, Dennis Carlson, has been sanctioned previously” for violating a
319. See, e.g., Timmins, supra note 92, at 1523 (noting that courts have held
that risk of harm through future disclosure is “entirely speculative”).
See, e.g., Parsons v. Gen. Motors Corp., 85 F.R.D. 724, 726 (N.D. Ga.
Beyond the harm accruing to a particular litigant from
relying on a later-modified protective order, some courts have
considered negative impacts on the system that might flow from
modifying an order. That was the scenario in Martindell, where
the Second Circuit required extraordinary circumstances to
modify an order to allow civil discovery to be used in a criminal
prosecution.321 Several witnesses relied on the protective order at
issue when deciding not to exercise their Fifth Amendment right
not to testify at a deposition.322
The court understandably worried that modifying the order
would deter future litigants in the same situation from
cooperating in civil discovery, particularly when criminal liability
loomed in the background.323 This reasoning doubtless has
common sense appeal—knowing that a protective order might be
modified to accommodate prosecutors would cause most people to
refuse to testify by invoking the Fifth.
The gap in the court’s reasoning, however, is twofold. First, it
is less than clear that future witnesses would know about the
previous protective order modification. Second, most cases do not
involve discovery that has a dual-purpose use as both civil
discovery and high-stakes criminal evidence, making the specter
of widespread discovery resistance unlikely in regular
circumstances.324 In the absence of empirical evidence to the
contrary, it seems unreasonable to assume that parties engaged
in ordinary civil litigation will resist discovery to a greater extent
than they would have anyway because of a possible
321. See Martindell v. Int’l Tel. & Tel. Corp., 594 F.2d 291, 296 (2d Cir.
1979) (discussing the reasons the “information in question” was not found to be
322. See id. at 295–96.
323. See id. (“In short, witnesses might be expected frequently to refuse to
testify pursuant to protective orders if their testimony were to be made
available to the Government for criminal investigatory purposes in disregard of
324. Cf. In re Agent Orange Prod. Liab. Litig., 821 F.2d 139, 147 (2d Cir.
1987) (finding that, in contrast to the deponents in Martindell, the proponents of
continued protection in this case “could not have relied on the permanence of the
protective order”), superseded by statute as stated in Iridium India Telecom Ltd.
v. Motorola, Inc., 165 F. App’x 878 (2d Cir. 2005).
The Wilk court appropriately created a presumption in favor
of modifying protective orders to give collateral litigants access to
discovery in the absence of substantial prejudice to the party
opposed to the modification.325 This modification-friendly
approach considers the same interests as the Martindell court but
accords them different importance.326
Instead of presuming the harm done to parties who allegedly
rely on later-modified protective orders, the Wilk court requires a
party resisting sharing via a modification to demonstrate actual
harm from reliance or other prejudice.327 At the same time, the
Wilk test accords proper weight to the benefits of shared
discovery by placing a thumb on that end of the scale. This
proofbased approach to reliance reasonably balances the interests at
A liberal modification regime also complements the current
proposed amendments to the civil rules. When litigants in
collateral cases modify a protective order and receive discovery
under the modification, the discovery process is aided in
collateral courts. Litigants receiving shared discovery information
are in a better position to understand the benefits of potential
discovery requests and to tailor them accordingly.328 Like any
325. See Wilk v. Am. Med. Ass’n, 635 F.2d 1295, 1299–1300 (7th Cir. 1980)
(explaining that the presumption for public access to pretrial discovery “should
operate with all the more force when litigants . . . use discovery in aid of
collateral litigation on similar issues” because “access in such cases materially
eases the tasks of courts . . . and speeds up . . . a lengthy process”), superseded
by rule as stated in Bond v. Utreras, 585 F.3d 1061 (7th Cir. 2009).
326. See State ex rel. Ford Motor Co. v. Manners, 239 S.W.3d 583, 587–88
(Mo. 2007). The Manners court has recognized a third strand, in addition to the
Martindell and Wilk tests, in the modification standard fight. See id. (“Arguably,
the standards recognize the same controlling criteria and differ only in the
weight that should be given to the criteria in their balancing.”).
327. See Wilk, 635 F.2d at 1301 (concluding that due to the “close similarity”
between the actions the “State is presumptively entitled to access to all of the
Wilk discovery on the same terms as the Wilk plaintiff”).
328. See FED. R. CIV. P. 26(b)(2)(C)(iii) (“[T]he burden or expense of the
proposed discovery outweighs its likely benefit, considering the needs of the
case, the amount in controversy, the parties’ resources, the importance of the
issues at stake in the action, and the importance of the discovery in resolving
the issues.”); see also COMMITTEE REPORT, supra note 6, at 80 (showing
committee’s suggested amendments to Fed. R. Civ. P. 26 regarding
proportionality in discovery).
form of discovery sharing, sharing through protective-order
modifications has the potential to reduce oft-repeating discovery
litigation in repeating cases.
C. Proportionality and Sharing
Although sharing does work in tandem with proportionality
to streamline discovery, the relationship has its limits. For
instance, shared discovery does not provide an independent
ground for courts and litigants to employ the proportionality
concept to limit discovery in most instances. At first glance,
shared discovery from a previous or contemporaneous case might
potentially be a “more convenient, less burdensome” source of
information.329 Moreover, shared discovery from another case
might reduce the “importance of the discovery in resolving the
issues.”330 Shared discovery in the possession of a requesting
party almost certainly does not preclude similar discovery in the
party’s own case based on these provisions in their current or
Despite the intuitive appeal of the argument, no
commentators and few courts have directly addressed the
question. Courts have, however, regularly addressed the closely
related question of whether litigants may request information
from another party when the requestor already has the
information in her possession.331 And they have routinely allowed
For example, courts should seldom disallow the deposition of
a witness in a case just because that witness has previously been
deposed in another case, even though that deposition has been
shared.333 The re-deposition of the witness should be allowed
329. FED. R. CIV. P. 26(b)(2)(C)(i).
330. FED. R. CIV. P. 26(b)(2)(C)(iii).
331. See, e.g., Wilson Land Corp. v. Smith Barney, Inc., 48 Fed. R. Serv. 3d
1297, 1302 (E.D.N.C. 2000) (“The mere fact that the matters regarding which
discovery is sought happen to be within the knowledge of the moving party is
not usually grounds for objection . . . .”).
332. Id.; Weiner v. Bache Halsey Stuart, Inc., 76 F.R.D. 624, 625 (S.D. Fla.
333. Cf. Hoh Co. v. Travelers Indem. Co., No. 87-0274 RCL, 1991 WL
because, particularly with party witnesses, litigants in the
current case are entitled to know that the party’s contentions at
deposition are based on current information for the current
case.334 Indeed, discovery is not limited to just an investigation of
facts; another important purpose of discovery is to ascertain the
parties’ contentions about the facts, claims, and defenses.335 Over
time, parties can subtly shift their positions or give a different
version of the facts. Memories change.
Different advocates advancing different claims ask different
questions. Forcing litigants to rely on potentially outdated
depositions to ascertain the current case’s factual and legal
posture would undermine the quality of fact finding at trial.
Thus, there is substantial necessity for the deposition in the
second case, and even though there is substantial subject-matter
overlap, the deposition is not, in any way that is cognizable by
Rule 26, “duplicative” or “cumulative” or even “available” from a
less costly source.
The same logic should also apply when the form of discovery
is dependent on written questions and a potentially variable
party response.336 Over time, circumstances, along with claims
and defenses, change. Litigants should have the benefit of their
own questions and the responses that flow from them. Requests
for admissions provide an apt example. Asking someone to admit
a fact or position presumes that the person asking has some
previous knowledge on which to base the question.337 Even if that
229948, at *3 (D.D.C. Oct. 25, 1991) (holding re-deposition allowed in light of
changed legal circumstances after remand).
334. See id. (stating that the proposition that a court may not “‘prohibit the
taking of a deposition altogether . . . absent extraordinary circumstances’”
applies “even when the party seeking discovery wishes to take a second
deposition of the same individual”).
335. See 8 WRIGHT ET AL., supra note 160, § 2014 (observing that the rule
prohibiting discovery of matters within knowledge of requesting party “could
hardly apply to discovery under the federal rules, since the purpose of the
discovery rules is not only to elicit unknown facts, but also to narrow and define
336. See, e.g., United States v. 58.16 Acres of Land, 66 F.R.D. 570, 573 (E.D.
Ill. 1975) (holding that interrogatories are proper even when inquiring party
already has knowledge).
337. See 8 WRIGHT ET AL., supra note 160, § 2014 (“[T]he very existence of a
request–for–admission procedure, which often implies some knowledge on the
knowledge flows from shared discovery, courts should not
preclude discovery as disproportionate.
Proportionality may, however, provide a mechanism to
preclude re-discovery of information in some very limited
circumstances. For instance, courts may decide that identical
requests, including requests to which responses would not change
from case to case, are “duplicative” or “cumulative” or simply not
necessary in light of the burden of producing information that
overlaps with already shared discovery.338 But this is a principle
of limited applicability.
Seldom will requests or responses be identical, or virtually
so, even in similar cases. The facts of each case, and the parties’
contentions and shading of the facts in each case, will almost
necessarily vary. The particular benefits of discovery for the
current case must, in most cases, be weighed anew against the
burdens of production despite the existence of shared responses
from a previous case.339 Thus, while sharing supports
proportionality through the better exercise of attorney discretion,
the practice is not an independent ground to limit discovery
under the current or amended proportionality rules.
Instead of providing a basis to forbid discovery, sharing
allows litigants to more carefully and thoughtfully employ
proportional discovery requests. Instead of searching in the dark
with unfocused discovery requests, attorneys can target
particularly beneficial information based on their knowledge of
other discovery productions in similar cases. This increased
accuracy reduces both the volume of discovery requests and
potential litigation about those requests stemming from
wellfounded or unfounded objections. In many instances, faced with
part of the requesting party of the matters covered by the requests for
admissions, shows that knowledge is no bar to use of the tools authorized by
Rules 26 to 36.”).
338. 6 MOORE ET AL., supra note 136, ¶ 26.41 (“A court will nevertheless
limit discovery of matters known to the discovering party when the discovery
appears to serve no purpose . . . .”).
339. See Wiwa v. Shell Petrol. Dev. Co. of Nigeria Ltd., 335 F. App’x 81, 84
(2d Cir. 2009) (holding that the district court committed clear error by granting
a Rule 26(b)(2)(C)(iii) request to limit discovery because of availability of
discovery in related cases when previous discovery was inadequate to address
issues in present case).
the choice between drafting requests for information already in
their possession (and potentially litigating objections to those
requests), litigants may well exercise their discretion to rely on
A. Pretrial Protective Orders .......................................2188 B. Fault Lines in the Protective-Order Debate ...........2193 C. Sharing Discovery in Similar Cases........................ 2198 2 . Modification .......................................................2209 III. Trends in Discovery Reform ..........................................2214 44 . See, e.g., Brown Bag Software v . Symantec Corp., 960 F.2d 1465 , 1472 ( 9th Cir . 1992 ) (finding a protective order that denied access to party's in-house counsel was not an abuse of discretion).