SSRN Id3069819

________________________________________________________________________
THE UNIVERSITY OF TEXAS

SCHOOL OF LAW
Law and Economics Research Paper
November 2017
Published originally in the Boston University Law Review at
97 B.U. L. REV. 1775 (2017)
Troll Check? A Proposal for

Administrative Review of Patent Litigation
Lauren Cohen
Harvard Business School & National Bureau of Economic Research
John M. Golden
The University of Texas School of Law
Umit G. Gurun
University of Texas at Dallas School of Management
Scott Duke Kominers
Harvard University
This paper can be downloaded without charge from the
Social Science Research Network at
http://ssrn.com/abstract=3069819.
________________________________________________________________________
Electronic copy available at: https://ssrn.com/abstract=3069819

TROLL CHECK? A PROPOSAL FOR
ADMINISTRATIVE REVIEW OF PATENT LITIGATION
LAUREN COHEN, JOHN M. GOLDEN, UMIT G. GURUN

& SCOTT DUKE KOMINERS
INTRODUCTION ............................................................................................. 1777

I. PRIMER ON PATENTS AND USPTO REVIEW ...................................... 1783
A. Patent Rights and Their Enforcement ........................................ 1783
B. The Patent Document and USPTO Review ............................... 1785
II. THE PATENT LITIGATION LANDSCAPE .............................................. 1790
A. Recipe for Trouble ..................................................................... 1791
1. Litigation Complexity, Longevity, and Cost ....................... 1791
2. Rise in Litigation, Especially Involving Software and
PAEs .................................................................................... 1794
3. Forum Shopping and Selling ............................................... 1797
L.E. Simmons Professor, Harvard Business School; Research Associate, National
Bureau of Economic Research. For helpful comments, the authors thank Alex Albright,
Jordan Barry, Jeremy Bock, Bob Bone, Julie Cohen, Chris Cotropia, Paul Courant, Mechele
Dickerson, Rebecca Eisenberg, Ken Felter, Roger Ford, Mark Gergen, John William Hatfield,
John Eric Humphries, Louis Kaplow, Bill Lee, Mark Lemley, Jonathan Masur, Scott
McCown, Michael Meurer, Martha Minow, Robbie Minton, Linda Mullenix, Justin Nelson,
Randall Rader, Michael Risch, Karen Sandrik, Chester Shiu, Ted Sichelman, Daniel Sokol,
Holger Spamann, Neel Sukhatme, David Taylor, Abe Wickelgren, and participants in
meetings, sessions, or workshops of the American Economics Association, the American Law
and Economics Association, the European Policy for Intellectual Property Association, the
Society for Institutional and Organizational Economics, PatCon 6, the Works in Progress
Intellectual Property Colloquium, the Georgetown University Law Center, the Tel Aviv
University Buchmann Faculty of Law, the Tilburg Law and Economics Center, the University
of Michigan Law School, and the University of Texas School of Law. The authors greatly
appreciate the support of National Science Foundation grants CCF-1216095, SciSIP-
1535813, and SES-1459912, as well as the Harvard Milton Fund and the Ng Fund of the
Harvard Center of Mathematical Sciences and Applications.
Loomer Family Professor, University of Texas School of Law.
Professor of Accounting and Finance, University of Texas at Dallas School of
Management; Research Economist, National Bureau of Economic Research.

MBA Class of 1960 Associate Professor, Harvard Business School; Faculty Affiliate,
Department of Economics, Harvard University; Affiliate, Harvard Center of Mathematical

Sciences and Applications; Associate, Harvard Center for Research on Computation and
Society; Research Economist, National Bureau of Economic Research.
1775

1776 BOSTON UNIVERSITY LAW REVIEW [Vol. 97:1775
B. Social Welfare Concerns with Patent Litigation ....................... 1801

1. General Social Welfare Concerns........................................ 1801
2. Concerns with PAEs ............................................................ 1804
III. PROPOSAL FOR ADMINISTRATIVE REVIEW ........................................ 1808
A. Economic Analysis of Administrative Review.......................... 1809
1. Basic Economic Theory ...................................................... 1809
2. A First Economic Model for Administrative Review.......... 1810
3. A Second, Calibrated Model for Administrative Review .... 1813
a. Model Structure ............................................................ 1813
b. Illustrative Results ........................................................ 1816
c. Analysis of Results ....................................................... 1821
i. PLRB Screening of Cases with Settlement ............ 1822
ii. PLRB Screening of Cases Without Settlement ...... 1823
d. General Welfare and Policy Implications ..................... 1824
B. Framework for Administrative Review ..................................... 1825
1. Proceedings Before the PLRB ............................................. 1825
2. Reporting Responsibilities and Sunset Provision ................ 1830
3. Agency Location for the PLRB ........................................... 1831
4. Defusing the Flood of Claims Concern............................ 1833
C. Complement or Substitute for Other Reforms ........................... 1836
CONCLUSION................................................................................................. 1841
The patent system is commonly justified as a way to promote social welfare

and, more specifically, technological progress. For years, however, there has
been concern that patent litigation is undermining, rather than furthering, these
goals. Particularly in the United States, the time, cost, and complications of
patent suits provide openings for opportunistic assertions of infringement.
This Article proposes a way to address information problems that facilitate
opportunistic assertion: an automatic process of administrative review at the
threshold of infringement lawsuits in U.S. district courts. The results of this
review would be non-binding but admissible in later court proceedings. Whether
conducted by an independent Patent Litigation Review Board or a division of
the U.S. Patent and Trademark Office, such review would: (1) help discourage
or bring to an earlier and less costly endrelatively weak patent-infringement
lawsuits; (2) strengthen the litigation and bargaining positions of patentees with
especially robust cases; (3) flag weaknesses in litigation positions to the benefit
of private parties and the courts; and (4) provide policymakers with information
that facilitates evaluation and adjustment of patent system performance. This
Article uses multiple economic models to show the likely benefits of early-stage
administrative review. Nonetheless, because of the fluid and complex nature of
the patent litigation landscape, this Article proposes that the review process
initially be adopted on a pilot basis.

2017] TROLL CHECK? 1777
INTRODUCTION
Patent litigation reform is coming. Just as in the lead up to the 2011 America
Invents Act (AIA),1 the United States Congress has entertained a host of patent
reform bills over a series of years.2 Many have focused on patent litigation.3 The
House of Representatives passed one of these litigation reform bills by a 325-91
vote in 2013,4 and supermajorities of the House and Senate Judiciary
Committees approved reform bills in 2015.5 Policymakers have promised a
renewed push for reform in 2017.6 Meanwhile, outside pressure for reform has
come not only from the usual suspects among industry stakeholders7 but also
from the popular press. In December 2013, the New York Times editorial board
cheered congressional consideration of sound proposals to restrict abusive
patent litigation.8 In August 2015, the Economist made patent reform its cover
story9 and came close to advocating abolition.10
1 See Joe Matal, A Guide to the Legislative History of the America Invents Act: Part I of
II, 21 FED. CIR. B.J. 435, 438-47 (2012) (chronicling the more than six years of legislative
activity that culminated in the AIA).
2 See infra notes 11-15 and accompanying text (describing recent congressional proposals
for patent reform).

3 See infra notes 11-15 and accompanying text.
4 The Innovation Act, HOUSE JUDICIARY COMMITTEE https://judiciary.house.gov/the-
innovation-act/ [https://perma.cc/XG8J-N69X] (last visited Sept. 18, 2017).

5 On June 4, 2015, the Senate Judiciary Committee voted 12-4 in favor of the Protecting
American Talent and Entrepreneurship Act (the PATENT Act), S. 1137. Kevin E. Noonan,
Senate Judiciary Committee Passes PATENT Act, JD SUPRA (June 5, 2015),
http://www.jdsupra.com/legalnews/senate-judiciary-committee-passes-85294/
[https://perma.cc/3HNQ-RYL3]. On June 11, 2015, the House Judiciary Committee voted 24-
8 in favor of the Innovation Act, H.R. 9. The Innovation Act, supra note 4.
6 Joseph Marks, Times Too Short for Patent Venue Revamp This Congress, Sponsor Looks
to Next Year, 92 PAT., TRADEMARK & COPYRIGHT J. 1108, 1108 (2016) (reporting that Senator
Jeff Flake had high hopes for [a patent venue reform bills] success in a new Congress, with
a new administration).
7 Cf. Greg Reilly, Linking Patent Reform and Civil Litigation Reform, 47 LOY. U. CHI. L.J.
179, 238 (2015) ([I]t is hard to dispute that procedural reform [of civil litigation] is inevitable
given the political influence of the large corporate interests most burdened . . . . (emphasis
omitted)).
8 Editorial, Congress Takes on Abusive Patent Suits, N.Y. TIMES (Nov. 30, 2013),
http://www.nytimes.com/2013/12/01/opinion/sunday/congress-takes-on-abusive-patent-
suits.html.
9 See ECONOMIST, Aug. 8, 2015 (showcasing a cover with the heading Set innovation
free! and subheading Time to fix the patent system).

10 Time to Fix Patents, ECONOMIST, Aug. 8, 2015, at 11 (One radical answer would be to
abolish patents altogether . . . .); cf. Intellectual Property: A Question of Utility, ECONOMIST,
Aug. 8, 2015, at 50, 52 ([A]ny lawmaker brave enough to propose doing away with [patents]
altogether . . . would face an onslaught from the intellectual-property lobby.).
Reform proposals have tended toward the dramatic. Some proposals have
threatened a revolution in patent litigationfor example, by proposing general
adoption of regular attorney fee shifting along a European loser pays model11
as opposed to the typical U.S. practice of shifting fees only rarely.12 Other
proposals have focused on disempowering so-called patent trolls13a
disparaging moniker for patent-assertion entities (PAEs) that specialize in the
ownership, licensing, and enforcement of patent rights.14 Although the reform
bills endorsed by the House and Senate Judiciary Committees in 2015 were
generally more modest than their predecessors, they were still draconian by the
usual standards of U.S. litigation reform.15
These reform efforts reflect concern that patent assertion activity is
undermining patent laws purpose to promote technological progress
specifically and social welfare more generally.16 A high overall caseload, high
litigation costs, and rampant forum shopping feed these concerns. Even after a
drop in district court filings in 2016, patent-suit filings are proceeding at about
double the rate of the year 2000.17 Moreover, these suits are not cheap. High-
stakes patent litigation tends to cost each side millions of dollars in attorney fees,
and even litigation in which less than $1 million is at stake tends to cost each
11 Adam Smith, Note, Patent TrollsAn Overview of Proposed Legislation and a Solution
That Benefits Small Businesses and Entrepreneurs, 9 OHIO ST. ENTREPRENEURIAL BUS. L.J.
201, 218 (2014) (observing that a congressional bill would essentially switch[] the traditional
assumption of American legal jurisprudence that parties pay their own attorney fees).
12 Cf. 35 U.S.C. 285 (2012) (providing that [t]he court in exceptional [patent-
infringement] cases may award reasonable attorney fees to the prevailing party).
13 See Smith, supra note 11, at 217 (discussing a bill ostensibly designed to target patent
trolls).
14 Cf. John M. Golden, Patent Trolls and Patent Remedies, 85 TEX. L. REV. 2111, 2112
(2007) (noting criticism of the patent trollapparently one of a class of patent owners who
do not provide end products or services themselves, but who do demand royalties as a price
for authorizing the work of others).
15 Compare Protecting American Talent and Entrepreneurship Act, S. 1137, 114th Cong.
3-7 (2015) (proposing patent litigation reforms), and Innovation Act, H.R. 9, 114th Cong.
3 (2015) (same), and Prachi Agarwal, Patent Troll: The Brewing Storm of Patent Reform in
the United States of America, 15 J. MARSHALL REV. INTELL. PROP. L. 63, 76-81 (2015)
(describing provisions of the Innovation and PATENT Acts), with Jared A. Smith & Nicholas
R. Transier, Trolling for an NPE Solution, 7 HASTINGS SCI. & TECH. L.J. 215, 238-39 (2015)
(listing various bills with direct and indirect anti-troll provisions).
16 See, e.g., Agarwal, supra note 15, at 64 ([P]atent trolls stifle, discourage, and threaten
innovation.); Smith, supra note 11, at 201 (noting that the rise of certain patent-assertion
entities . . . has renewed discussion . . . about the state and effectiveness of the current patent
law).
17 According to Lex Machina, over four thousand five hundred patent suits were filed in
2016 compared to just over two thousand three hundred in 2000. See LEX MACHINA,
https://lexmachina.com/ (last visited Aug. 31, 2017).
side several hundred thousand dollars.18 Further, the concentration of new suits
in just two of the nations ninety-four federal judicial districts,19 the Eastern
District of Texas and the District of Delaware, has become astounding.20
According to a representative tally, in each year from 2012 through 2016, as
well as in the first several months of 2017, over 40% of all new patent suits have
been filed in these two districts.21 A May 2017 decision by the Supreme Court
on patent venue22 will most likely shuffle the deck of suit locations but appears
unlikely to fully address concerns with forum shopping. In the immediate
aftermath of the decision, filings in the Eastern District of Texas predictably
decreased, but filings in the District of Delaware (also predictably23) increased,
with the result being that the cumulative percentage of filings in the two districts
remained at about 40% of the total.24
In recent years, multiple tweaks to patent law have responded to concerns
about patent assertion and litigation. Courts have taken a more restrictive
approach to granting injunctions against adjudged infringers, thereby curtailing
the ability of patent holders, particularly PAEs, to extract exorbitant licensing
fees by threatening to shut down a factory or line of business.25 Courts have also
18 AM. INTELLECTUAL PROP. LAW ASSN, 2015 REPORT OF THE ECONOMIC SURVEY 37
(2015) [hereinafter AIPLA 2015 SURVEY] (reporting median costs to pursue a patent suit to
completion of $600,000 when less than $1 million is at stake, $2 million when between $1
million and $10 million is at stake, and over $3 million when more than $10 million is at
stake).
19 Court Role and Structure, U.S. COURTS, http://www.uscourts.gov/about-federal-
courts/court-role-and-structure [https://perma.cc/Z63M-5EPU] (last visited Sept. 18, 2017).

20 U.S. GOVT ACCOUNTABILITY OFFICE, GAO-16-490, INTELLECTUAL PROPERTY: PATENT
OFFICE SHOULD DEFINE QUALITY, REASSESS INCENTIVES, AND IMPROVE CLARITY 14-16 (2016)
(observing that, in 2015, nearly 50% of all patent infringement defendants were named in
cases filed in the Eastern District of Texas); id. at 17 fig.3 (indicating that Delaware is the
second most popular district in which to bring a patent infringement suit).
21 This information was obtained using Lex Machinas search functionality. See LEX
MACHINA, supra note 17.

22 TC Heartland LLC v. Kraft Food Brands Grp. LLC, 137 S. Ct. 1514, 1521 (2017) (As
applied to domestic corporations, reside[nce] in [the patent venue statute] refers only to the
State of incorporation. (first alteration in original)).
23 See infra text accompanying note 166.
24 See Matthew Bultman, Gilstraps Venue Test an Encouraging Sign for Patent Owners,
LAW360 (July 6, 2017, 9:37 PM), https://www.law360.com/articles/941916/gilstrap-s-venue-

test-an-encouraging-sign-for-patent-owners (reporting that, in the first thirty-eight days after
the Supreme Courts decision in TC Heartland, sixty-one patent cases, equal to 14% of all
new cases, were filed in the Eastern District of Texas and 117 cases were filed in the District
of Delaware).
25 See eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 394 (2006) (rejecting the
general rule in favor of granting injunctions against adjudged infringers); Thomas F. Cotter
& John M. Golden, Empirical Studies Relating to PatentsRemedies, in RESEARCH
tightened the standards for awarding reasonable royalty damages26 and have
made the shifting of attorney fees more likely in light of meritless positions in
litigation.27 Other court decisions have strengthened patentability requirements
of subject-matter eligibility and nonobviousness, thereby making many suits
more likely to fail in response to a motion for dismissal or summary judgment.28
Additional reforms have been more purely procedural. Various district courts
have adopted local rules specific to patent cases that, in principle, should speed
up and streamline litigation but might have contributed to growth in case
filings.29 Through the AIA, Congress expanded opportunities for post-issuance
review of patent validity by the U.S. Patent and Trademark Office (USPTO).30
Congress also restricted joinder in patent cases in a way intended to reduce the
number of defendants sued by PAEs.31 In 2015, amendments to the Federal
Rules of Civil Procedure abrogated a model form for pleading that had enabled
patent holders to file complaints featuring little more . . . than the name and
number of the patent and an allegation of infringement.32
HANDBOOK ON THE ECONOMICS OF INTELLECTUAL PROPERTY LAW (Peter Menell, David

Schwartz & Ben Depoorter eds., forthcoming 2018) (describing studies indicating that
injunctions have become rare in patent-infringement suits brought by PAEs), available at
https://ssrn.com/abstract=2665680 [https://perma.cc/LM8M-JRMG].
26 John M. Golden, Patent Privateers: Private Enforcements Historical Survivors, 26
HARV. J.L. & TECH. 545, 605 (2013) (collecting cases).

27 See Octane Fitness, LLC v. ICON Health & Fitness, Inc., 134 S. Ct. 1749, 1754 (2014)
(abrogating a Federal Circuit rule requiring both subjective bad faith and objective[]
baseless[ness] for a court to award attorney fees for pursuing a weak litigation position).
28 Cf. id.; Golden, supra note 26, at 605 (mentioning Supreme Court decisions that
tightened the requirement of patentable subject matter . . . and the requirement of

nonobviousness).
29 Golden, supra note 26, at 607 (Like highway improvements that attract too many
drivers and make traffic congestion worse, litigation reforms can aggravate, rather than
alleviate, tendencies toward excessive litigation.).
30 John M. Golden, Working Without Chevron: The PTO as Prime Mover, 65 DUKE L.J.
1657, 1666 (2016) (The AIAs provisions for inter partes, post-grant, and covered business
method review notably empower the [USPTOs Patent Trial and Appeal Board (PTAB)] to
conduct more trial-like proceedings than those previously conducted by the [US]PTO.).
31 In response to complaints about PAEs simultaneously suing numerous unrelated
accused infringers in inconvenient venues, Congress overrode the Federal Rules of Civil
Procedures liberal approach to joinder in favor of patent-specific rules that generate a greater
prospect of litigating the same factual and legal questions numerous times. David O. Taylor,
Patent Misjoinder, 88 N.Y.U. L. REV. 652, 654-55 (2013).
32 Matthew Bultman, Stricter Pleading Requirements Take Effect Dec. 1, LAW360 (Nov.
30, 2015, 3:55 PM), https://www.law360.com/articles/717900/stricter-patent-pleading-

requirements-take-effect-dec-1; cf. Patricia W. Hatamyar Moore, The Anti-Plaintiff Pending
Amendments to the Federal Rules of Civil Procedure and the Pro-Defendant Composition of
the Federal Rulemaking Committees, 83 U. CIN. L. REV. 1083, 1125-26 (2015) (criticizing
Nonetheless, the flow of patent disputes into the district courts and USPTO
post-issuance proceedings remains high,33 and complaints about abusive patent
litigation continue seemingly unabated.34 This should not be a surprise. The
tweaks to the patent system have not altered three fundamental structural
problems. First, each year the USPTO continues to receive hundreds of
thousands of patent applications and to issue hundreds of thousands of patents.35
At these rates, one cannot reasonably expect the USPTO to perform more than
a relatively cursory examination of patents before they issue.36 Although the
USPTO contributes its own partial corrective by processing about two thousand
petitions for post-issuance review each year,37 these post-issuance efforts still
leave much cleanup work on patent validity to private parties and the courts.
Second, the USPTO contributes even less to the resolution of questions about
patent scope and infringement than it does to the resolution of questions about
patent validity. Indeed, even in adversarial post-issuance proceedings, the
USPTO does not address questions of patent infringement.38 Moreover, in both
pre-issuance and post-issuance proceedings, the USPTO uses a different
[t]he pending abrogation of Rule 84 and all thirty-six of the official forms following the
[Federal Rules of Civil Procedure]).
33 See infra notes 279-81 and accompanying text (describing how, despite various patent
reforms, the filing rates for new patent suits and USPTO post-issuance proceedings remain
high).
34 See, e.g., Impact of Bad Patents on American Businesses Before the Subcomm. on
Courts, Intellectual Property and the Internet of the H. Comm. on the Judiciary, 115th Cong.
3 (2017) (statement of Sean Reilly, Senior Vice President and Associate General Counsel,
The Clearing House Payments Company L.L.C., also Clearing House Association L.L.C.,
also Financial Services Roundtable) (expressing concern about the ability of patent holders to
extract settlement payments in meritless litigation).
35 U.S. PATENT & TRADEMARK OFFICE, U.S. PATENT ACTIVITY: CALENDAR YEARS 1790 TO
THE PRESENT (2016) (reporting that, in every year from 2006 through 2015, the USPTO
received over four hundred thousand utility patent applications and issued over 150,000 utility
patents and over twenty thousand design patents).
36 See infra notes 73-76 and accompanying text (discussing the time available for review
of a typical patent application).

37 See U.S. PATENT & TRADEMARK OFFICE, PATENT TRIAL AND APPEAL BOARD STATISTICS
3 (2016) [hereinafter PTAB STATISTICS 2016] (reporting the filing of between one thousand
seven hundred and one thousand nune hundred petitions for inter partes, covered business
method, or post-grant review in each of fiscal years 2015 and 2016); U.S. PATENT &
TRADEMARK OFFICE, EX PARTE REEXAMINATION FILING DATA (2016) (reporting between two
hundred and two hundred fifty filings for ex parte reexamination in each of fiscal years 2015
and 2016).
38 John M. Golden, Patentable Subject Matter and Institutional Choice, 89 TEX. L. REV.
1041, 1053 (2011) ([T]he USPTO has historically had no direct involvement with
determinations of whether an accused infringers conduct in fact constitutes
infringement . . . .).
standard for claim construction than that used by the district courts.39 Third, the
several hundred thousand to several million dollar price of patent litigation40
creates real concerns about nuisance suits and access to justice.41 The cost barrier
of patent litigation can be especially problematic for small businesses, which are
generally unable to appear in court pro se.42
Substantially effective patent reform needs to address one or more of these
fundamental problems. This Article addresses the fundamentals by proposing a
new administrative filter for patent suits that, in its complete form, would apply
automatically to all patent suits filed in the district courts. Unlike the USPTO,
this Articles proposed Patent Litigation Review Board (PLRB) would review
questions of claim construction, infringement, and unenforceability, as well as
validity, and it would use a claim-construction standard identical to that of the
courts. Further, the PLRB would focus not on providing final decisions on a
complete record, but on providing non-binding assessments of whether, even at
a preliminary, pre-discovery stage, there is clear evidence that a party to the case
should prevail on one or more issues.43
The details of the proposed framework for PLRB review are intended to be
measured, but this would not be a minor reform. The institution of PLRB review
would constitute the most significant institutional change to the patent system
since 1982, when patent appeals were centralized under a new United States
Court of Appeals for the Federal Circuit.44 The general imposition of PLRB
39 Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2144-45 (2016) (upholding a
USPTO regulation prescribing the broadest reasonable construction standard).

40 See supra note 18 and accompanying text (describing typical costs of patent litigation).
41 FED. TRADE COMMN, PATENT ASSERTION ENTITY ACTIVITY: AN FTC STUDY 4 (2016)
(observing that the typical license royalties earned by a class of PAEs that typically sued
potential licensees and settled shortly afterward were less than $300,000, an amount
approximat[ing] the lower bound of early-stage litigation costs).
42 John M. Golden, Litigation in the Middle: The Context of Patent-Infringement
Injunctions, 92 TEX. L. REV. 2075, 2093 (2014) (U.S. courts have ruled that business entities
generally cannot be represented pro se . . . .).
43 A previous proposal for mandatory reexamination of patents asserted in litigation
contemplated USPTO reexamination of a limited subset of validity issues, rather than non-
binding, preliminary review for any potential question of patent validity, enforceability, or
infringement. See Benjamin J. Bradford & Sandra J. Durkin, A Proposal for Mandatory
Patent Reexaminations, 52 IDEA 135, 164-65 (2012) (noting the currently limited scope of
reexaminations while proposing possible extension to encompass additional sources of prior
art and challenges based on a charge of inadequate patent disclosure or patent claim
indefiniteness).
44 Cf. John M. Golden, The Federal Circuit and the D.C. Circuit: Comparative Trials of
Two Semi-Specialized Courts, 78 GEO. WASH. L. REV. 553, 555 (2010) (describing the
creation of the Federal Circuit as an even more recent and radical experiment in semi-
specialization than the preceding creation of the D.C. Circuit).
review would provide a systematic backup to USPTO review and a substantially

more accessible and informative front end for district court litigation.
Although PLRB judgments would not be substantively binding on courts,
they, and the opinions behind them, would be admissible in court, and parties
and the courts would likely give them substantial weight.45 As a result, this
Article contends that PLRB review would accomplish the following: (1) help
discourageor bring to an earlier and less costly endrelatively weak patent-
infringement lawsuits; (2) strengthen the hands of patentees with especially
robust cases; (3) flag weaknesses in litigation positions to the benefit of both
private parties and the courts; and (4) provide policymakers with more readily
aggregated information that facilitates evaluation and adjustment of patent
system performance.
The Article proceeds in three parts. Part I provides a primer on patents and
existing processes of administrative review at the USPTO. Part II describes
aspects of the current landscape of patent litigation in the United States. Part III
presents multiple economic models and eight tables illustrating the expected
positive effects and practicability of PLRB review. Part III also describes details
of a suggested framework for administrative review, including a proposal that
such review initially be adopted on a pilot basis. Finally, Part III discusses how
the proposed framework operates as an alternative or complement to other
potential or already implemented adjustments of the patent system.
I. PRIMER ON PATENTS AND USPTO REVIEW

To motivate this Articles proposal for a new process of administrative
litigation review, this Part provides a brief discussion of the basic nature of
patent rights and current processes for USPTO review.
A. Patent Rights and Their Enforcement

Patents provide their owners with territorially limited rights to exclude others
from the making, use, sale or offer for sale, or importation of covered subject
matter.46 Under the current standard patent term, these rights last from the time
a patent issues until twenty years from the first relevant filing of an application
with the USPTO or a qualifying foreign patent office.47 For a party to be liable
45 Cf. Irina D. Manta, Bearing Down on Trademark Bullies, 22 FORDHAM INTELL. PROP.
MEDIA & ENT. L.J. 853, 867 (2012) (proposing USPTO review of trademark cease-and-desist
letters).
46 See 35 U.S.C. 154(a)(1) (2012) (describing the patent owners right to exclude); id.
271(a) (listing acts that constitute patent infringement).

47 See id. 154(a)(2) (specifying the default term for new U.S. patents).
for patent infringement, that party need not know of the infringed patent.48 Nor
need the party have derived the covered subject matter in any way from the
patents inventors or owners.49 Although knowledge of a relevant patent is
generally required for liability for indirect infringement in the nature of aiding
and abetting,50 direct infringement by engaging in such acts as manufacture, use,
sale, or importation occurs regardless of whether any of the parties have
knowledge of the pertinent patent and regardless of the fact that the manufacturer
independently developed all the relevant technology.51 Hence, if a consumer
uses in the United States a smartphone, which was purchased from a retail store
in the United States and imported by a manufacturer who independently
developed all the technology associated with the smartphone, the consumer,
retail store, and manufacturer could all be liable for direct infringement of a U.S.
patent covering technology in the smartphone.
When a patent owner suspects that its patent is being infringed, the owner can
sue in district court52 for relief such as lost profits or reasonable royalty
damages,53 enhanced damages,54 or an injunction.55 The patent owner who
brings such a suit need not be the inventor of the patented technology or the
owner of the relevant rights at the time the patent issued. By statute, patent rights
may be assigned to others.56 This assignability permits the sale of patent rights
to entities that specialize in the acquisition and assertion of patent rights without
any direct involvement in the development, sale, or use of the covered subject
48 Commil USA, LLC v. Cisco Sys., Inc., 135 S. Ct. 1920, 1926 (2015) (observing that
[d]irect infringement is a strict-liability offense for which a defendants mental state is

irrelevant).
49 See John M. Golden, Principles for Patent Remedies, 88 TEX. L. REV. 505, 515 (2010)
([U]nlike copyright infringement, patent infringement does not require copying and, as a
general rule, does not excuse independent creation.).
50 Global-Tech Appliances, Inc. v. SEB S.A., 563 U.S. 754, 765 (2011) (concluding that
knowledge of the relevant patent is required for indirect infringement).

51 Commil, 135 S. Ct. at 1926.
52 See 28 U.S.C. 1338(a) (2012) (setting forth the district courts original jurisdiction
over patent cases); 35 U.S.C. 281 (A patentee shall have remedy by civil action for
infringement of his
patent.).
53 See 35 U.S.C. 284 (providing for compensatory damages for patent infringement);
JANICE MUELLER, PATENT LAW 321 (4th ed. 2013) (describing lost profits and reasonable
royalty as the two primary analytical methods of computing compensatory damages for
patent infringement).
54 35 U.S.C. 284 ([T]he court may increase the damages up to three times the amount
found or assessed.).
55 Id. 283 (granting courts the power to grant injunctions in accordance with the
principles of equity to prevent the violation of any right secured by patent).

56 Id. 261 (providing that patents, or any interest therein, shall be assignable in law by
an instrument in writing).
matter itself, entities variously called patent aggregators, patent assertion

entities, non-practicing entities (NPEs),57 or patent trolls.58
B. The Patent Document and USPTO Review

The patent document is the primary indicator of the scope of associated patent
rights.59 This document consists of drawings, a written description of the alleged
invention, and patent claims that are drafted by the patent applicant or its agents,
submitted to the USPTO, and commonly amended during the examination
process.60 The written description is required to disclose the alleged invention
and the manner and process of making and using it in a manner sufficient to
show that the inventor was in possession of the invention at the time of filing
a patent application,61 and to enable any person skilled in the art to which it
pertains, or with which it is most nearly connected, to make and use the same.62
Patent claims are numbered clauses at the end of the patent document63 that are
required to particularly point[] out and distinctly claim[] the subject matter
which the inventor or a joint inventor regards as the invention.64 The claims are
the primary reference points for determining the scope of rights under a
57 NPEs, which are distinguished from practicing entities (PEs), are patent holders that
do not themselves produce the sort of saleable products, services, or components thereof that
infringement suits tend to target. Mark P. Gergen, John M. Golden & Henry E. Smith, The
Supreme Courts Accidental Revolution? The Test for Permanent Injunctions, 112 COLUM. L.
REV. 203, 244 (2012).
58 LANNING G. BRYER, SCOTT J. LEBSON & MATTHEW D. ASBELL, INTELLECTUAL PROPERTY
STRATEGIES FOR THE 21ST CENTURY CORPORATION: A SHIFT IN STRATEGIC AND FINANCIAL
MANAGEMENT 145 (2011) (listing various monikers for so-called [p]atent trolls).
59 See Phillips v. AWH Corp., 415 F.3d 1303, 1313 (2005) (en banc) (Importantly, the
person of ordinary skill in the art is deemed to read the [patent] claim term not only in the
context of the particular claim in which the disputed term appears, but in the context of the
entire patent, including the specification.).
60 See ROBERT PATRICK MERGES & JOHN FITZGERALD DUFFY, PATENT LAW AND POLICY:
CASES AND MATERIALS 13-26 (6th ed. 2013) (describing the patent document and its relation
to the application).
61 Ariad Pharms., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010) (en banc)
(holding that 112 of the Patent Act contains a written description requirement demanding
that the disclosure of the [patent] application relied upon reasonably conveys to those skilled
in the art that the inventor had possession of the claimed subject matter as of the filing date).
62 35 U.S.C. 112(a) (2012).
63 John M. Golden, Construing Patent Claims According to Their Interpretive
Community: A Call for an Attorney-Plus-Artisan Perspective, 21 HARV. J.L. & TECH. 321,
322 (2008).
64 35 U.S.C. 112(b).
particular patent.65 Under the doctrine of equivalents, however, there is often

room for a patent to cover matter substantially equivalent to what is claimed
even though not within the claims literal scope.66
The USPTO subjects each patent application to substantive review by one or
more examiners.67 To be validly patented, an invention must: (1) satisfy
requirements that the invention claimed by the application have at least minimal
functionality;68 (2) be novel and nonobvious at the relevant time to one skilled
in the relevant art;69 (3) be adequately described by the patent document;70 and
(4) be delineated in a way that inform[s] those skilled in the art about the scope
of the invention with reasonable certainty.71
Examiners generally have expertise associated with the subject matter they
review,72 but the quality of their examination suffers from severe constraints. In
large part because the USPTO receives hundreds of thousands of patent
applications each year,73 examiners have only very limited time to review
individual applications. Even if one heroically assumes that each of the
USPTOs approximately nine thousand examiners74 spends two thousand hours
per year examining applications, one ends up with an estimate of thirty hours for
an individual examiner to review each of the roughly six hundred thousand new
65 See Golden, supra note 63, at 322 (Claimsnumbered clauses at the end of a patent
are meant to provide notice of what a patent covers and to describe a patented invention in a
way that distinguishes it from prior art.).
66 See Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 732 (2002).
67 See F. Scott Kieff, The Case for Registering Patents and the Law and Economics of
Present Patent-Obtaining Rules, 45 B.C. L. REV. 55, 70-71 (2003) (noting that the
examination process is also called patent prosecution).
68 See MUELLER, supra note 53, at 321 ([T]he substantive threshold for satisfying the
utility requirement is relatively low.).

69 35 U.S.C. 101-103 (setting out requirements of novelty and nonobviousness).
70 See supra text accompanying notes 61-62 (discussing U.S. patent laws written
description and enablement requirements).

71 Nautilus, Inc. v. Biosig Instruments, Inc., 134 S. Ct. 2120, 2139 (2014) (discussing the
requirement of patent claim definiteness).

72 See F. SCOTT KIEFF ET AL., PRINCIPLES OF PATENT LAW 99 (6th ed. 2013) (When an
application reaches an examining group, it is assigned to the appropriate art (i.e., technology)
unit and then to a particular examiner.); MERGES & DUFFY, supra note 60, at 52 (noting the
specialization of the examiners, who are assigned to a particular technology).
73 See U.S. PATENT & TRADEMARK OFFICE, U.S. PATENT STATISTICS CHART: CALENDAR
YEARS 1963-2015 (2017) [hereinafter CALENDAR YEAR PATENT STATISTICS],

http://www.uspto.gov/web/offices/ac/ido/oeip/taf/us_stat.htm [https://perma.cc/B3GG-
YUSL].
74 U.S. PATENT & TRADEMARK OFFICE, FISCAL YEAR 2015 PERFORMANCE AND
ACCOUNTABILITY REPORT 14 (2015), http://www.uspto.gov/sites/default/files/documents/

USPTOFY15PAR.pdf [https://perma.cc/GN8B-XDCV] (indicating that the USPTO
employed 9161 patent examiners at the end of fiscal year 2015).
applications filed per year.75 In these thirty hours, the examiner must: (1) read
the application and understand its technical subject matter; (2) search and review
antecedent material (prior art) that could indicate that the claimed invention
lacks novelty or nonobviousness; (3) evaluate satisfaction of patentability
requirements; (4) write up any relied-upon bases for rejecting the application;
and (5) possibly engage in telephone or in-person interviews with the applicant
or the applicants agents.76
Limits on examiner time suffice to indicate that the USPTOs pre-issuance
review can act only as a rough screen for patent application quality. But there
are other reasons to suspect that the USPTO issues a large number of patents or,
at least, individual patent claims that do not really satisfy requirements for
patentability.77 Not only are examiners limited in the time that they can search
prior art, they are generally limited in their ability to consult outside experts78
and other sources of information, including the Internet, during the eighteen
months that applications typically remain confidential.79 Moreover, examiners
bear the burden of proof. From the moment of a patent applications filing, an
entitlement to an issued patent is effectively presumed: the examiner must show
75 See CALENDAR YEAR PATENT STATISTICS, supra note 73 (showing that USPTO received
589,410 utility patent applications, 39,097 design patent applications, and 1140 plant patent
applications in 2015); cf. John M. Golden, Proliferating Patents and Patent Laws Cost
Disease, 51 HOUS. L. REV. 455, 496-97 (2013). More realistic estimates of examiner time
per application put the average time available for these activities at about twenty hours per
application, rather than thirty. See Chris J. Katopis, Perfect Happiness?: Game Theory as a
Tool for Enhancing Patent Quality, 10 YALE J.L. & TECH. 360, 373 (2008) (It is estimated
that, on average, an examiner must examine eighty-seven applications per year, spending
approximately nineteen hours on each application.); Mark A. Lemley, Rational Ignorance
at the Patent Office, 95 NW. U. L. REV. 1495, 1500 (2001) (The total average time the
examiner spends on all these tasks over the two- to three-year prosecution of the patent is
eighteen hours. (emphasis omitted)).
76 See KIEFF ET AL., supra note 72, at 99-102, 101 nn.38-43; Lemley, supra note 75,
at 1500.
77 See Joseph Farrell & Robert P. Merges, Incentives to Challenge and Defend Patents:
Why Litigation Wont Reliably Fix Patent Office Errors and Why Administrative Patent
Review Might Help, 19 BERKELEY TECH. L.J. 943, 944-45 (2004) (discussing grounds for
believing that the USPTO issues many patents that should not be enforced, either on
economic or on legal grounds).
78 Cf. Doug Lichtman & Mark A. Lemley, Rethinking Patent Laws Presumption of
Validity, 60 STAN. L. REV. 45, 62 (2007) (advocating a review process enabling examiners
not only to spend at least one full month researching each purported invention, but also to
hire relevant outside experts).
79 See MPEP 904.02(c), at 900-44 (rev. 9th ed. Nov. 2015) (stating that examiner Internet
use must comply with confidentiality requirements); Golden, supra note 63, at 336
([E]xaminers face tight restrictions on their ability to consult any outside evidence, never
mind outside experts.).
non-patentability by a preponderance of evidence.80 The USPTO might further

tilt the balance against rejection through the agencys openly declared concern
for fee-paying customers, such as patent applicants, and, more particularly,
through the agencys traditional assignment of performance credit for an
examiners closing of a case through the grant of patent rights.81
Given the deficiencies of the USPTOs pre-issuance review of patents, the
patent system unsurprisingly provides opportunities for private-party challenges
to the validity of an issued patent. Most notably for our purposes, a party sued
for infringement or confronting an immediate threat of suit for infringement82
has long been able to challenge the validity of the relevant patent in district
court.83 A party making such a challenge must prove invalidity by clear and
convincing evidence,84 but success in such challenges is far from rare. In cases
in which questions of novelty or nonobviousness are litigated to a final
judgment, challengers win about half the time.85 Selection effectsproducts of
parties presumed selectivity in determining which issues are litigated to final
judgment as opposed to settled, voluntarily dismissed, or never even asserted
mean that such litigation-based invalidation rates do not provide a great
indication of the underlying percentage of issued patent claims that are invalid.86
Regardless, it seems accepted that an accused infringer can generally mount a
substantial validity challenge to at least some of a patents claims.87
In the early 1980s, Congress began responding to uncertainty about issued
patent claims validity by establishing administrative post-issuance proceedings
80 Sean B. Seymore, The Presumption of Patentability, 97 MINN. L. REV. 990, 1023 (2013)
(noting an examiners dual burden of building a prima facie case of [non]patentability and
carrying the ultimate burden of persuasion).
81 Patrick A. Doody, How to Eliminate the Backlog at the Patent Office, 37 AIPLA Q.J.
395, 411-13 (2009).

82 See, e.g., Arkema Inc. v. Honeywell Intl, Inc., 706 F.3d 1351, 1357 (Fed. Cir. 2013);
KIMBERLY A. MOORE, PAUL R. MICHEL & TIMOTHY R. HOLBROOK, PATENT LITIGATION AND
STRATEGY 50 (3d ed. 2008) (Declaratory judgment actions can be a sword for the alleged
infringer as well as a shield.).
83 See 35 U.S.C. 282(b) (2012) (identifying potential defenses in any action involving
the validity or infringement of a patent).

84 Microsoft Corp. v. i4i Ltd. Pship, 564 U.S. 91, 95 (2011) (holding that the Patent Act
requires an invalidity defense to be proved by clear and convincing evidence).

85 See Michael D. Frakes & Melissa F. Wasserman, Does the U.S. Patent and Trademark
Office Grant Too Many Bad Patents? Evidence from a Quasi-Experiment, 67 STAN. L. REV.
613, 621 (2015) (noting the frequently cited statistic that courts invalidate nearly half of all
litigated patents that make it to final judgment).
86 See id. at 621-22 (observing that litigated patents are a highly select sample of patents
whose characteristics vary substantially from allowed patents in general).

87 See Mark A. Lemley & Carl Shapiro, Probabilistic Patents, 19 J. ECON. PERSP. 75, 76
(The risk that a patent will be declared invalid is substantial.).

through which the validity of patent claims might be challenged or clarified.88

The available types of such proceedings and the frequency of their overall use
have grown over the past three decades.89 Now there are four such types, with
hundreds of proceedings launched each year.90 The four types are: (1) ex parte
reexaminations to evaluate new questions of novelty or nonobviousness based
on prior-art patents or printed publications;91 (2) inter partes review
proceedings, in which a private party can effectively litigate novelty or
nonobviousness based on the same art forms;92 (3) post-grant review
proceedings in which a party can litigate essentially any kind of validity question
as long as the request for review comes within nine months of the relevant
patents issuance;93 and (4) covered business method review proceedings whose
procedure and permissible substance track those of post-grant review.94
Although post-issuance USPTO proceedings have become a booming
business,95 substantial statutory limitations on these proceedings mean they
cannot completely substitute for district court litigation. As noted above, ex parte
reexamination and inter partes review are restricted to a limited subset of
potential bases for challenging a patent claims validitynamely, arguments of
lack of novelty or nonobviousness relative to previously available patents or
88 Gregory Dolin, Dubious Patent Reform, 56 B.C. L. REV. 881, 899 (2015) (noting that
reexamination processes established in 1981 and 1999 were intended to increase the
reliability of issued patents).
89 Id. at 883-84 (discussing the increase in post-issuance review mechanisms beginning in
1981).
90 See U.S. PATENT & TRADEMARK OFFICE, PATENT TRIAL AND APPEAL BOARD STATISTICS
7 (2015) [hereinafter PTAB STATISTICS 2015], http://www.uspto.gov/sites/default/files/

documents/2015-11-30%20PTAB.pdf [https://perma.cc/DP97-S389] (reporting PTAB
institution of about nine hundred inter partes review, post-grant review, and covered business
method review proceedings in fiscal year 2015, not including joinders); U.S. PATENT &
TRADEMARK OFFICE, PERFORMANCE AND ACCOUNTABILITY REPORT FISCAL YEAR 2015, at 196
tbl.14A (2015) https://www.uspto.gov/sites/default/files/documents/USPTOFY15PAR.pdf
[https://perma.cc/64KG-42AU] (reporting grants of about two hundred requests for ex parte
reexamination in fiscal year 2015).
91 35 U.S.C. 303 (2012); MERGES & DUFFY, supra note 60, at 1040.
92 See 35 U.S.C. 311; MERGES & DUFFY, supra note 60, at 1050-51.
93 See MERGES & DUFFY, supra note 60, at 1047.
94 See Eric C. Cohen, A Primer on Inter Partes Review, Covered Business Method Review,
and Post-Grant Review Before the Patent Trial and Appeal Board, 24 FED. CIR. B.J. 1, 22
(2014) (The transitional proceeding [for covered business method patents] employs the
standards of post-grant review, except that there is no nine-month deadline for filing and the
estoppel provisions of post-grant review do not apply to civil actions.).
95 See Golden, supra note 30, at 1667 (From mid-2014 through the third quarter of 2015,
filings for inter partes post-issuance proceedings before the PTAB arrived at a rate of about
one hundred fifty per month.).
printed publications.96 Post-grant review and covered business method review

enable challenges to validity on any grounds but are available only within nine
months of a patents issuance or for covered business method patents,
respectively.97
Two more concerns with USPTO proceedings are that they are either non-
adversarial or expensive. Ex parte USPTO proceedings like pre-issuance
examination and post-issuance reexamination generally exclude continuing
input from patent challengers and are commonly viewed as too likely to favor
patentees.98 The three adversarial forms of post-issuance proceedings are
significantly less costly than district court proceedings but still pricey for many
potential litigants. Between USPTO fees and fees for attorneys and experts, a
party to inter partes review is likely to spend $200,000 by the end of motion
practice and about $350,000 overall.99
This Article proposes a new type of administrative proceeding that is expected
to be cheaper for the parties involved and offers an adversarial administrative
forum to hear disputes over patent claim construction and infringement as well
as patent validity. Part II builds the initial case for this new type of administrative
proceeding by describing the troubled nature of the current patent litigation
landscape.
II. THE PATENT LITIGATION LANDSCAPE

Many critics of the patent system believe patent litigation produces
problematic incentives in general or at least in circumstances involving PAEs.100
Patent litigation can hurt more than help overall incentives for innovation if
litigation outcomes improperly allocate economic rewards or if any benefits
from improved allocation through litigation are outweighed by social costs of
96 See supra text accompanying notes 91-92.

98 See Dale L. Carlson & Robert A. Migliorini, Patent Reform at the Crossroads:
Experience in the Far East with Oppositions Suggests an Alternative Approach for the United
States, 7 N.C. J.L. & TECH. 261, 270 (2006) (describing ex parte reexamination as unduly
favorable to the patentee because of the very limited involvement of the third-party
requester).
99 See AIPLA 2015 SURVEY, supra note 18, at 38.
100 See, e.g., James Bessen, Jennifer Ford & Michael Meurer, The Private and Social Costs
of Patent Trolls, 34 REGULATION 26, 26 (concluding that lawsuits brought by NPEs

substantially reduce [technology companies] incentives to innovate); Herbert Hovenkamp,
Antitrust and the Patent System: A Reexamination, 76 OHIO ST. L.J. 467, 559-60 (2015)
(describing the recent sharp increase in PAE activity as a ballooning crisis in the patent
system).
litigation itself.101 This Part describes reasons for concern that patent litigation
has become a morass that often undermines the objectives that Congress intends
patents to serve.
A. Recipe for Trouble

A patent-infringement suit in district court is a form of complex litigation that
typically features technical subject matter, multiple stages, and high costs.102
The Judicial Conference of the United States has attested to the burden that
patent-infringement suits impose on courts by assigning these suits the fourth
highest case weight for civil suits in district courts, trailing only death-penalty
habeas, environmental, and civil RICO cases.103 Three aspects of the patent
litigation landscape are particularly worth highlighting (1) the complexity,
longevity, and cost of litigation; (2) concerns with PAEs and software patents;
and (3) the apparent prevalence of forum shopping.
1. Litigation Complexity, Longevity, and Cost

Patent litigation in the district courts and other forms of complex civil
litigation share many common characteristics. After a patent holder files suit and
the defendant answers and potentially countersues, [t]he parties proceed to fact
and expert discovery, motion practice, pretrial briefing, and trial.104
But patent litigation in the district courts typically features a relatively distinct
claim construction phase in which a judge determines the meaning of contested
patent claim language.105 The claim construction phase of patent cases ordinarily
precedes summary judgment filings and occurs after much, if not all,
discovery.106 In a conventional version of this phase, the parties brief disputed
claim terms107 and provide a technology tutorial,108 and the trial judge holds an
101 See Golden, supra note 49, at 517-18 (discussing costs imposed by the patent system
and noting that excessive rewards from patent rights could induc[e] the diversion of
resources from more socially productive activity).
102 See infra text accompanying notes 103-31.
103 FED. JUDICIAL CTR., 2003-2004 DISTRICT COURT CASE-WEIGHTING STUDY: FINAL
REPORT TO THE SUBCOMMITTEE ON JUDICIAL STATISTICS OF THE COMMITTEE ON JUDICIAL

RESOURCES OF THE JUDICIAL CONFERENCE OF THE UNITED STATES 5 tbl.1 (2005).
104 PETER S. MENELL ET AL., PATENT CASE MANAGEMENT JUDICIAL GUIDE 2-4 (3d ed.
2016).
105 See id. at 5-3 (describing claim construction as one of the most distinctive aspects of
patent litigation).
106 See id. at 5-5 to 5-6 (discussing practices with respect to discovery both before and after
claim construction).
107 Ronald J. Schutz & Jonathan D. Goins, Case Management Issues in Patent Litigation,
5 SEDONA CONF. J. 1, 2 (2004) (Markman briefs and hearings are a critical part of patent
litigation proceedings . . . .).
108 MENELL ET AL., supra note 104, at 5-15.
oral hearing and issues a claim construction orderoften called a Markman

orderthat interprets relevant terms.109 Claim construction can significantly
clarify the strength of parties positions110 but often comes only after much time
and money has already been expended. According to a study using data from
Lex Machina, for the approximately 10% of patent suits initiated and terminated
between 2000 and 2010 that resulted in a claim-construction order, the average
time from case filing to claim construction was 1.8 years.111
A variety of litigation phases can follow claim construction. As indicated
above, the district courts claim construction is often followed by a summary
judgment phase, in which parties file, support, and dispute motions for summary
judgment.112 If a case is not resolved by summary judgment or by settlement
before or after rulings on summary judgment, the case proceeds with further
pretrial developments such as the drafting of jury instructions.113 Courts have
recognized a right to a jury trial in patent cases involving a claim for damages,114
and most present-day trials occur before a jury.115 These jury trials may be
followed by post-trial motions for a new trial or judgment as a matter of law.116
The district courts might conduct additional post-jury-verdict proceedings on
attorney fee shifting, enhanced damages, or injunctive relief.117 After a district
courts final judgment, a party may appeal to the U.S. Court of Appeals for the
109 See id. at 2-4 to 2-5; Pauline M. Pelletier, The Impact of Local Patent Rules on Rate
and Timing of Case Resolution Relative to Claim Construction: An Empirical Study of the
Past Decade, 8 J. BUS. & TECH. L. 451, 467 (2013) (noting that the Markman order is so
called after the seminal case on claim construction).
110 See Schutz & Goins, supra note 107, at 2 ([T]he courts rulings on claim construction
and interpretation often determine the outcome of the case.).

111 Pelletier, supra note 109, at 477 (describing results from a study of 28,377 patent
cases).
112 See MENELL ET AL., supra note 104, at 6-10.
113 Id. at 7-2.
114 Devon Curtis Beane, Note, Whose Right Is It Anyway?: The Evisceration of an
Infringers Seventh Amendment Right in Patent Litigation, 2011 U. ILL. L. REV. 1853, 1858
(noting courts differential treatment of cases where plaintiffs seek damages and those
where the patentee seeks only injunctive relief).
115 Mark A. Lemley, Why Do Juries Decide If Patents Are Valid?, 99 VA. L. REV. 1673,
1674 & n.1 (2013) (Lawyers, scholars, and judges take for granted that when a patent case
goes to trial, that trial will almost always be before a jury.).
117 Id.
Federal Circuit in Washington, D.C.118 Appeals frequently result in cases being

remanded for further proceedings.119
Key aspects of this process of multistage litigation are that it tends to take
years and that at least half of its overall cost tends to occur during discovery
phases that precede any trial.120 Even before any appeal, district court
proceedings that run through trial commonly span about two years, with even a
district known for quick case schedules, the Eastern District of Texas, having
a median time to trial of 1.8 years during the period from 2000 to 2007.121
Moreover, the costs of patent litigation are typically high. Patent litigation has
commonly been called a sport of kings, the sense being that it is typically so
expensive that only extraordinarily well-heeled plaintiffs and defendants can
afford to pursue it.122 Much of the cost is associated with the process of
discovery, in which sides frequently exchange huge quantities of documents
relating to the claimed invention, the nature of the accused product or process,
and the developmental histories of both.123 Testifying and non-testifying experts
are commonly employed to analyze and explain aspects of such material,124 and
the two sides can incur additional expenses developing instructional or
illustrative graphics to help render relevant technology comprehensible to
generalist judges and juries.125 According to the results of a biannual survey by
the American Intellectual Property Lawyers Association, such activities lead to
litigation costs per side that tend to rise with perceived stakes and that often total
millions of dollars.126
Of course, the true costs of litigation likely exceed out-of-pocket costs, and
these overall costs might be even more heavily weighted toward pretrial
118 See John M. Golden, The Supreme Court as Prime Percolator: A Prescription for
Appellate Review of Questions in Patent Law, 56 UCLA L. REV. 657, 664 (2009) (describing
the patent jurisdiction of the Federal Circuit); John M. Golden, Response, Too Human?
Personal Relationships and Appellate Review, 94 TEX. L. REV. 70, 77 (2016) (discussing the
location and national jurisdiction of the Federal Circuit).
120 See infra note 123.
121 Daniel Klerman & Greg Reilly, Forum Selling, 89 S. CAL. L. REV. 241, 265 & n.131
(2016) (Nationwide, the median time to trial in [patent cases from 2000 to 2007] was two
years.).
122 Golden, supra note 42, at 2077 & n.15.
123 See AIPLA 2015 SURVEY, supra note 18, at 37-39 (listing median litigation costs
through discovery that generally either approximately equal or exceed half of median total
litigation costs).
124 Edward G. Poplawski, Selection and Use of Experts in Patent Cases, 27 AIPLA Q.J. 1,
3 (1999) ([E]xpert testimony is virtually essential in assisting the trier of fact to understand
the evidence and to resolve factual issues in [patent] litigation.).
125 See Lisa C. Wood, Making Your Case with Graphics: An Interview, 22 ANTITRUST 103,
104-05 (2008).
126 See supra note 18 and accompanying text.
discovery. A defendants business might operate under a cloud of uncertainty127

until it can implement a design-around of asserted patent rights, which is a
redesign of the defendants products or processes that the defendant can claim
steers well clear of any charges of infringement.128 Further, patent litigation can
be most disruptive to a defendants business during discovery, in substantial part
because the number of key employees subjected to questioning, depositions, and
document requests during this stage can far exceed the number of employees
called as witnesses at trial.129
In sum, patent litigation in the district courts tends to be an expensive,
multistage process that, in the absence of settlement, takes years to conclude.
These aspects of patent litigation can have negative social effects, including the
relative chilling of innovative activity that might result either from undue
barriers to the enforcement of patent rights130 or from the encouragement of so-
called nuisance suits focused on extracting settlement payments by threatening
innovators with litigation costs.131
2. Rise in Litigation, Especially Involving Software and PAEs

Concerns about the complexity, longevity, and cost of patent litigation have
not stifled its growth. Patent suit filings have increased dramatically since the
year 2000, with the number of new suits rising from about two thousand three
hundred to over four thousand five hundred annually.132 Although part of this
127 Martin J. Adelman & Gary L. Francione, The Doctrine of Equivalents in Patent Law:
Questions that Pennwalt Did Not Answer, 137 U. PA. L. REV. 673, 682 (1989) ([U]ncertainty
about the scope of patent protection hinders both patent holders and potential defendants from
assessing the possible outcome of litigation or from making other business decisions . . . .
(footnote omitted)).
128 Cf. KIEFF ET AL., supra note 72, at 68.
129 Cf. Colleen Chien, Startups and Patent Trolls, 17 STAN. TECH. L. REV. 461, 472 (2014)
(stating that surveyed entrepreneurial companies commonly reported that resolving [a

patent] demand required founder time (73%) and distracted from the core business (89%)
(footnote omitted)).
130 See D. Rosenberg & S. Shavell, A Model in Which Suits Are Brought for Their Nuisance
Value, 5 INTL REV. L. & ECON. 3, 3 (1985) (stating that a plaintiff could be likely to
prevail . . . but . . . still not want to go to trial because the litigation costs would exceed the
expected judgment).
131 See Sergio J. Campos, Christopher S. Cotton & Cheng Li, Deterrence Effects Under
Twombly: On the Costs of Increasing Pleading Standards in Litigation, 44 INTL REV. L. &
ECON. 61, 67 (2015) (describing nuisance lawsuits as frivolous lawsuits intended to entice
settlement from a likely innocent defendant who wants to avoid litigation costs).
132 LEX MACHINA, supra note 17; see also Lauren Cohen, Umit G. Gurun & Scott Duke
Kominers, The Growing Problem of Patent Trolling: Cash-Hungry Patent Trolls Are
Squelching Innovationand Should Be Screened Out, 352 SCIENCE 521, 521 (2016)
[hereinafter Cohen et al., Growing Problem] (The last decade has seen a sharp rise in patent
litigation in the United States . . . .); Lauren Cohen, Umit Gurun & Scott Duke Kominers,
rise reflects multiplication of suits as a result of the new joinder restrictions

enacted in 2011,133 the dramatic rise in patent-suit defendants provides a less
ambiguous indication of the growing intensity of patent litigation. According to
one estimate, the number of defendants in new patent cases increased from about
six thousand five hundred or less in each of the three years from 1999 through
2001 to about twelve thousand or more in each of the three years from 2012
through 2014.134 Two focal points of concern with such growth in litigation have
been software patent cases and lawsuits brought by PAEs.
In 2013, the Government Accountability Office (the GAO) reported that
software-related patents were at issue in nearly half of the patent-infringement
suits filed from 2007 through 2011, with suits over software-related patents
accounting for nearly two-thirds of defendants in new patent-infringement suits
and just under 90% of the increase in defendants over this period.135 The GAO
also found that suits involving software-related patents had an unusual tendency
to persist before the courts, thereby presumably running up higher litigation
costs. Specifically, the GAO found a statistically significant difference between
suits involving software-related patents, of which 82% settled compared with
89% of suits that did not involve software-related patents.136
Concern about software patent litigation has extended beyond the GAO.
Commentators have argued that software-related patents tend to raise particular
problems for assessments of patent scope, the validity of patent claims, and the
proper value of patent damages.137 Moreover, a recent Supreme Court decision
has cast doubt on the subject-matter eligibilityand thus validityof many
Patent Trolls: Evidence from Targeted Firms 5 (Natl Bureau of Econ. Research, Working
Paper No. 20322, 2014) [hereinafter Cohen et al., Patent Trolls], http://www.nber.org/
papers/w20322.pdf [https://perma.cc/7XE7-DDDV] (The amount of patent-related litigation
has increased tenfold since 2000.).
133 See supra text accompanying note 31 (describing change in joinder rules); see also
Christopher A. Cotropia, Jay P. Kesan & David L. Schwartz, Unpacking Patent Assertion
Entities (PAEs), 99 MINN. L. REV. 649, 655 (2014) ([M]ost of the differences between the
years [2010 and 2012] are likely explained by, and attributable to, a change in the joinder
rules adopted in 2011 . . . .).
134 Matthew Sag, IP Litigation in U.S. District Courts: 1994-2014, 101 IOWA L. REV. 1065,
1082 & tbl.3 (2016) (The real trend in patent litigation over the past two decades can be seen
in the number of defendants filed against.).
135 U.S. GOVT ACCOUNTABILITY OFFICE, GAO-13-465, INTELLECTUAL PROPERTY:
ASSESSING FACTORS THAT AFFECT PATENT INFRINGEMENT LITIGATION COULD HELP IMPROVE
PATENT QUALITY 21 (2013) [hereinafter 2013 GAO REPORT].
136 Id. at 25 n.46.
137 See, e.g., JAMES BESSEN & MICHAEL J. MEURER, PATENT FAILURE: HOW JUDGES,
BUREAUCRATS, AND LAWYERS PUT INNOVATORS AT RISK 187 (2008) (contending that, because
software is an abstract technology, software patents are especially likely to have unclear
boundaries and give rise to opportunistic litigation (emphasis omitted)).
software-related patent claims.138 As a result, general concerns with uncertainty

about patent quality and vagaries in patent litigation are commonly intensified
for software patents.139
Like software patents, PAEs loom large in the recent growth in patent
litigation. Evidence indicates that PAEs have driven most of that growth, with
the numbers of suits brought annually by entities that actively commercialize or
otherwise practice their patented inventions being comparatively flat.140
Evidence on whether overly broad or otherwise invalid (bad) patents
substantially explain growth in PAE activity is ambiguous, but recent large-
sample empirical evidence suggests that, on average, entities such as PAEs buy
and litigate lower quality patents. Initial studies based on relatively small
samples indicated that patents held or asserted by PAEs were of higher than
normal quality, in the sense that they were more highly cited and had wider
technical breadth than was typical of patents overall.141 Such small-sample
studies tend to be particularly liable to data-collection biases, however.142 Large-
sample empirical studies have provided significant indications of comparatively
low quality for PAE-owned patents. Specifically, such studies have indicated
PAE patent portfolios disproportionately comprised patents whose claims were
allowed by patent examiners who spent relatively less time reviewing and
138 Alice Corp. v. CLS Bank Intl, 134 S. Ct. 2347, 2352 (2014) (holding that merely
requiring generic computer implementation fails to transform [a specified] abstract idea into
a patent-eligible invention).
139 See 2013 GAO REPORT, supra note 135, at 45 (suggesting that software-related patents
might be a better focus of patent reform efforts than any particular class of patent owners, and
also noting that most of the suits brought by [patent monetization entities from 2007 through
2011] involved software-related patents).
140 Lauren Cohen, Umit G. Gurun & Scott Duke Kominers, Empirical Evidence on the
Behavior and Impact of Patent Trolls: A Survey, in PATENT ASSERTION ENTITIES AND
COMPETITION POLICY 27, 28 fig.3.1 (D. Daniel Sokol ed., 2017) [hereinafter Cohen et al.,
Empirical Evidence].
141 See, e.g., Timo Fischer & Joachim Henkel, Patent Trolls on Markets for Technology
An Emperical Analysis of NPEs Patent Acquisitions, 41 RES. POLY 1519, 1526 (2012)
(NPEs acquire patents that, on average, lie in denser technology fields, received more
forward citations, have more claims, are older, and lie in more crowded technology fields than
patents acquired by practicing firms.); Michael Risch, Patent Troll Myths, 42 SETON HALL
L. REV. 457, 481 (2012) ([T]raditional patent quality measures imply at the very least that
NPE patents look a lot like other litigated patents. If one believes that these measures indicate
patent quality, then NPE patents would appear to be of equal or higher quality.); Sannu K.
Shrestha, Note, Trolls or Market-Makers? An Empirical Analysis of Nonpracticing Entities,
110 COLUM. L. REV. 114, 146 (2010) (citing data indicating that the NPE patents in the
sample have disproportionately higher values than randomly selected litigated patents and the
analyzed set of litigated peer patents).
142 See Cohen et al., Empirical Evidence, supra note 140, at 35 (noting concerns about the
representativeness of the small samples used in earlier studies).

narrowing claims,143 PAEs were significantly more likely than practicing

entities (PEs)144 to have patent claims invalidated,145 and PAEs tended to
litigate patents that were older and closer to expiration than those litigated by
PEs.146
In short, there is substantial evidence that not only the increased volume of
patent litigation but also the emergence of software patents and PAEs as major
players in patent litigation might be straining the patent systems ability to
advance social goals. An additional troublesome aspect of the current patent
litigation landscape, the apparent prevalence of forum shopping, arguably
exacerbates these strains.
3. Forum Shopping and Selling

From at least the early 1990s until the spring of 2017, a patent owner
commonly had a wide choice of fora in which to file suit. By act of Congress,
[a]ny civil action for patent infringement may be brought in the judicial district
where the defendant resides, or where the defendant has committed acts of
infringement and has a regular and established place of business.147 Congress
defines [r]esidency in another section of the same title of the U.S. Code.148
This definition provides, in part, that an entity with the capacity to sue and be
sued in its common name under applicable law . . . reside[s], if a defendant, in
any judicial district in which such defendant is subject to the courts personal
jurisdiction with respect to the civil action in question.149
143 Josh Feng & Xavier Jaravel, Who Feeds the Trolls?: Patent Trolls and the Patent
Examination Process 4 (July 12, 2016) (unpublished manuscript), http://ssrn.com/

abstract=2838017 [https://perma.cc/C35K-LTE6] (We find that patents purchased by NPEs
are, on average, granted by examiners who allow more incremental patents and patents with
vaguer language.).
144 See supra note 57 (implicitly defining PEs by comparison to NPEs).
145 John R. Allison, Mark A. Lemley & David L. Schwartz, How Often Do Non-Practicing
Entities Win Patent Suits? 40 (Stanford Law Sch. John M. Olin Program in Law & Econ.,
Paper No. 485, 2016), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2750128
[https://perma.cc/7SS2-5MUQ] (Operating companies won more often [in litigation] than
NPEs in our data set.).
146 Brian J. Love, An Empirical Study of Patent Litigation Timing: Could a Patent Term
Reduction Decimate Trolls Without Harming Innovators?, 161 U. PA. L. REV. 1309, 1312
(2013) (Product-producing companies predominantly enforce their patents soon after they
issue and complete their enforcement activities well before their patents expire. NPEs, on the
other hand, begin asserting their patents relatively late in the patent term and frequently
continue to litigate their patents to expiration.).
147 28 U.S.C. 1400(b) (2012).
148 Id. 1391(c).
149 Id. 1391(c)(2).

In 1990, the Federal Circuit held that this definition determines the scope of
residency for purposes of patent venue.150 Under this ruling, companies that
produced consumer products sold throughout the United States could often be
sued for patent infringement in any of the countrys ninety-four judicial
districts.151 A Supreme Court decision issued in May 2017 overruled the Federal
Circuit, however, holding that a domestic corporation resides only in its State
of incorporation for purposes of the patent venue statute.152 Partly because a
patent infringement suit may also be brought in the judicial district . . . where
the defendant has committed acts of infringement and has a regular and
established place of business,153 the practical significance of the Courts
holding on residency remains to be determined.154
In any event, for at least some patentees, there is an alternative forum beyond
the district courts. Upon complaint by a private party, the International Trade
Commission (the ITC), an independent agency created [t]o protect domestic
industry from unfair practices,155 can launch proceedings that culminate in an
order prohibiting the importation of specified articles or prohibiting domestic
activities involving imported matter.156 But the overall number of ITC
proceedings to enforce patents is much smaller than the number of district court
cases. Whereas several thousand patent suits are initiated annually in district
courts,157 the ITC lists less than eighty total Section 337 proceedings as having
been instituted in each fiscal year between 2010 and 2015 and estimates that less
150 VE Holding Corp. v. Johnson Gas Appliance Co., 917 F.2d 1574, 1584 (Fed. Cir. 1990)
(describing personal jurisdiction as the first test for venue), abrogated by TC Heartland
LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514 (2017); Elizabeth P. Offen-Brown,
Note, Forum Shopping and Venue Transfer in Patent Cases: Marshalls Response to TS Tech
and Genentech, 25 BERKELEY TECH. L.J. 61, 64-65 (2010) (discussing interpretation of 28
U.S.C. 1400(b)).
151 Jeanne C. Fromer, Patentography, 85 N.Y.U. L. REV. 1444, 1451 (2010) ([A] patent
plaintiff may frequently choose to initiate a lawsuit in virtually any federal district court.);
Mark Liang, The Aftermath of TS Tech: The End of Forum Shopping in Patent Litigation
and Implications for Non-Practicing Entities, 19 TEX. INTELL. PROP. L.J. 29, 39 (observing
that patentees commonly can sue in essentially any district because of the nationwide scope
of sales).
152 TC Heartland LLC, 137 S. Ct. at 1517.
153 28 U.S.C. 1400(b).
154 Raytheon Co. v. Cray, Inc., 2:15-cv-01554, 2017 WL 2813896, *12 (E.D. Tex. June
29, 2017) (concluding that a fixed physical location in the district is not a prerequisite to
proper venue based on the location of infringing acts and place of business).
155 Sapna Kumar, The Other Patent Agency: Congressional Regulation of the ITC, 61 FLA.
L. REV. 529, 544 (2009).

156 19 U.S.C. 1337(a)(1), (b)(1), (d)-(f) (2012); Cotter & Golden, supra note 25 (noting
the ITCs remedial powers).

157 See supra text accompanying note 132 (describing the volume of patent litigation in
district courts).
than fifty such proceedings will be instituted in each of fiscal years 2016 and
2017.158 Because ITC proceedings account for only a small fraction of patent-
infringement litigation, this Article focuses on suits filed in district courts.
There has been strong evidence of rampant forum shopping in the district
courts. Broadly based venue in patent cases enabled a remarkable concentration
of new patent suits in two federal districts commonly lying far from the centers
of infringers operations, the District of Delaware and the Eastern District of
Texas.159 According to data compiled by Lex Machina,160 more than 40% of the
thousands of new patent-infringement suits filed in each year from 2012 through
the first several months of 2017 were filed in one of these two districts.161 In
2015, nearly 45% of suits were filed in the Eastern District of Texas alone.162
Further, evidence suggests that, among patent holders, PAEs and other NPEs
were especially likely to sue in the Eastern District of Texas.163 Plaintiff-friendly
local procedural and administrative rules and realities, such as local rules that
expedite pretrial proceedings,164 comparatively predictable assignment of judges
158 U.S. INTL TRADE COMMN, ANNUAL PERFORMANCE PLAN, FY 2016-2017 AND ANNUAL
PERFORMANCE REPORT, FY 2015, at 83 tbl.C.2 (2016) https://www.usitc.gov/documents/
usitc_2016_2017_app_and_2015_apr.pdf [https://perma.cc/E36E-2MFZ] (outlining total
Section 337 proceedings instituted and completed for fiscal years 2010 through 2015 and
estimating total Section 337 proceedings expected to be instutituted and completed in fiscal
years 2016 and 2017).
159 See J. Jonas Anderson, Court Competition for Patent Cases, 163 U. PA. L. REV. 631,
632 (2015) ([N]early half of the six thousand patent cases filed in 2013 were filed in just two
of [the ninety-four districts in the United States]: the District of Delaware and the Eastern
District of Texas.); Klerman & Reilly, supra note 121, at 249 tbl.1 (listing the Eastern
District of Texas and the District of Delaware as the two most popular venues for patent cases
from 2007 through the first half of 2015).
160 Lex Machina makes data on intellectual property litigation in the United States freely
available to qualifying public interest users such as university scholars. Public Interest, LEX
MACHINA, https://lexmachina.com/public-interest/ [https://perma.cc/TCN4-4R8Q] (last
visited Sept. 18, 2017).
161 This information was obtained from a Lex Machina page entitled All Court Case
Filings by Year. See LEX MACHINA, supra note 17.

162 Id.
163 Cohen et al., Growing Problem, supra note 132, at 521 ([T]he preponderance of NPE
patent litigation . . . is brought in the Eastern District of Texas . . . .); cf. Allison, Lemley &
Schwartz, supra note 145, at 30 (finding that only 8.2% of suits brought by PEs and litigated
to judgment were filed in the Eastern District of Texas, whereas the analogous figure was
26.9% for litigated-to-judgment suits brought by NPEs).
164 Liang, supra note 151, at 43-46 (pointing to procedural factors such as filing-to-trial
time).
to individual cases,165 judicial reluctance to grant summary judgment,166 and the

nature of local jury pools,167 might explain much of the disproportionate
caseload of Delaware and the Eastern District of Texas. Given this backdrop, the
Supreme Courts holding on residence for patent venue purposes might do much
to shift cases from the Eastern District of Texas to the District of Delaware, the
leading state of incorporation for publicly traded companies.168 But such a shift
might fail to relieve concern about combined concentration of cases in these two
districts.169
Although some commentators defend forum shopping and district courts
self-differentiation as legitimate legal practices,170 the apparent intensity of the
phenomenon in patent law is at least ironic, just as the concern that judges might
be biasing procedures to attract categories of plaintiffs is troubling.171 Congress
specifically sought to limit forum differentiation in patent cases by creating the
Federal Circuit, an appellate court with nationwide jurisdiction over patent
165 Klerman & Reilly, supra note 121, at 254 (Patentees have the unique opportunity in
the Eastern District of Texas to choose their judge.); id. at 281 (observing that the District of
Delawares small sizefour allotted judgeships, with a vacancy for most of the past
decadeprovides significant judicial predictability).
166 Liang, supra note 151, at 43-46; see also Klerman & Reilly, supra note 121, at 251
([J]udges in the Eastern District of Texas grant summary judgment at less than one-quarter
the rate of judges in other districts.); id. at 281 (The District of Delaware has . . . a low
summary judgment rate.).
167 Klerman & Reilly, supra note 121, at 254 ([J]uries in the Eastern District [of Texas]
have a pro-patentee reputation.); Yan Leychkis, Of Fire Ants and Claim Construction: An
Empirical Study of the Meteoric Rise of the Eastern District of Texas as a Preeminent Forum
for Patent Litigation, 9 YALE J.L. & TECH. 193, 210-11 (2007) (reporting a 90% win rate for
patentees in jury trials in the Eastern District of Texas from 1998 to 2006, as opposed to the
national win rate of 68%).
168 Sean J. Griffith & Alexandra D. Lahav, The Market for Preclusion in Merger
Litigation, 66 VAND. L. REV. 1053, 1054-55 (2013) (More than half of all public companies
and over 60% of the Fortune 500 are incorporated in Delaware.).
170 See, e.g., Mary Garvey Algero, In Defense of Forum Shopping: A Realistic Look at
Selecting a Venue, 78 NEB. L. REV. 79, 111 (1999) (concluding that forum shopping should
be recognized as a legitimate practice); Debra Lyn Bassett, The Forum Game, 84 N.C. L.
REV. 333, 335 (2006) (contending that forum shopping is a legitimate, expressly authorized
action when more than one forum satisfies the requisite legal criteria).
171 See Anderson, supra note 159, at 634-35 (hypothesizing that forum shopping in patent
law is driven, at least in part, by federal district courts competing for litigants); Klerman &
Reilly, supra note 121, at 250 (contending that judges in the Eastern District [of Texas] have
consciously sought to attract patentees . . . by departing from mainstream doctrine in a variety
of procedural areas).
appeals.172 Evidence of trial-level forum shopping suggests that district courts

abilities to distinguish themselves through procedure and practice might threaten
a fundamental congressional goal.173
B. Social Welfare Concerns with Patent Litigation

How do problems with patent litigation affect social welfare? The patent
system is commonly justified on grounds that it promotes technological progress
or increases social welfare more generally.174 But there has long been concern
that patents, like other forms of intellectual property rights, can undermine these
aims by acting more as an innovation-sapping tax or rent-seekers delight than
as a beneficent stimulus for innovation.175 Such concern has become particularly
sharp with respect to modern patent litigation.
1. General Social Welfare Concerns

Two basic causes of social concern with patent litigation are its costs and the
vagaries of its outcomes. Section II.A.1 discussed the private costs of patent
litigation. This Section explores the social welfare implications of such costs and
of possibilities for error and bias in litigation outcomes.
As noted previously,176 high litigation costs can encourage so-called nuisance
or strike suits, lawsuits of relatively low legal merit whose expected value for
patent holders primarily results from anticipation of the accused infringers
willingness to pay to avoid expected litigation costs.177 Because of the
172
Markman v. Westview Instruments, Inc., 517 U.S. 370, 390 (1996) (It was just for the
sake of such desirable uniformity [in treatment of individual patents] that Congress created
the Court of Appeals for the Federal Circuit . . . .); Rochelle Cooper Dreyfuss, The Federal
Circuit: A Case Study in Specialized Courts, 64 N.Y.U. L. REV. 1, 7 (1989) (stating that
proponents of the Federal Circuits creation contended that channelling patent cases into a
single appellate forum would create a stable, uniform law and would eliminate forum
shopping).
173 See Megan M. La Belle, The Local Rules of Patent Procedure, 47 ARIZ. ST. L.J. 63, 65
(2015) (observing that, [w]ith the surge of patent litigation in recent years, district courts
around the country have crafted unique procedural rules for patent suits).
174 Golden, supra note 49, at 509 (describing the utilitarian goal that is standard in modern
accounts).
175 See, e.g., Fritz Machlup & Edith Penrose, The Patent Controversy in the Nineteenth
Century, 10 J. ECON. HIST. 1, 1 (1950).

176 See supra text accompanying notes 40-41, 131.
177 Colleen Chien et al., Santa Clara Best Practices in Patent Litigation Survey, 42 AIPLA
Q.J. 137, 180-81 (2014) (Nuisance suits are suits for patent infringement that the defendant
is motivated to settle due to the high cost of litigation, even if the patent is weak or its
economic value is low.); William H.J. Hubbard, A Fresh Look at Plausibility Pleading, 83 U.
CHI. L. REV. 693, 722 (2016) ([T]he defendants primary objective in a nuisance suit is
avoiding its litigation costs.).
disconnect between the private value of such litigation-cost-driven enforcement

activities and the social value of the patented invention, there is little reason to
believe that such activities will provide patent-holder awards well-tailored to
advance the patent systems social goals. Meanwhile, by acting as a tax on
litigation targets, nuisance suits and the threat of nuisance suits can discourage
socially valuable innovation by depressing the net private value of activities that
such suits target.178
Even if a private party is not sued, that party might incur costs in an effort to
limit the risk of a patent-infringement suit. For example, the party might
undertake an expensive design-around179 or stockpile patents defensively, either
to deprive others of patent-assertion opportunities or to be better positioned to
respond to a patent-infringement suit with a countersuit.180 Although such risk-
responsive behavior can generate social benefits through associated investments
in innovation or patenting, there seems to be little reason to believe that private
parties evasive or arms-race-style behavior will yield social benefits that
outweigh its social costs.181
On the patent enforcement side, anticipated litigation costs can act as a barrier
to use of the courts. Cash-strapped patent holders might find that their patents
are effectively unenforceable outside the slim prospect of their case being
cherry-picked for enforcement on a contingent-fee basis.182 Even capital-rich
patent holders might find that litigation costs effectively preclude enforcement
of patents of moderate value by causing meritorious suits to have a negative net
present value. As with nuisance suits effect on the value of the innovative
178 See Golden, supra note 49, at 517 & n.71 (Patents can impair dynamic efficiency by
impeding follow-on development . . . .); Rebecca S. Eisenberg, Noncompliance,
Nonenforcement, Nonproblem? Rethinking the Anticommons in Biomedical Research,
45 HOUS. L. REV. 1059, 1079 (2008) (discussing survey results on how concerns of patent
clearance can affect innovative activities).
179 See Carl Shapiro, Injunctions, Hold-Up, and Patent Royalties, 12 AM. L. & ECON. REV.
280, 290-99 (2010).

180 Michael Risch, Patent Portfolios as Securities, 63 DUKE L.J. 89, 100-01 (2013)
(reporting that motivations to acquire large portfolios of patents include preventing

assertion of those patents against the buyer and protecting against lawsuits filed by
competitors).
181 See John M. Golden, Injunctions as More (or Less) than Off Switches: Patent-
Infringement Injunctions Scope, 90 TEX. L. REV. 1399, 1409 (2012) (suggesting that design-
around activity might best be viewed as a means to mitigate the patent systems costs, rather
than as a primary justification for patents); Golden, supra note 14, at 2154 (observing that
patent dtente is neither costless nor uninterrupted and can involve use of patents to create
barriers to new entrants (footnote omitted)).
182 See David L. Schwartz, The Rise of Contingent Fee Representation in Patent Litigation,
64 ALA. L. REV. 335, 358 (2012) (reporting survey results indicating that contingent-fee
lawyers for patent cases decline two-thirds to over ninety-nine percent of potential cases).
activities of alleged infringers, litigation costs depression of patents effective

value seems likely to be substantially untethered to social interests.
Error and bias in patent litigation outcomes can also undermine patents
promise by inducing a misallocation of resources from more to less socially
productive activities. The generalist judges and juries of district courts might
commonly struggle with patent laws technical legal provisions or technological
and economic subject matter.183 Lack of mastery of the law or technological
subject matter could help to randomize case outcomes, leading to an unusually
high percentage of cases in which trial courts award victory to a party who would
not prevail under perfect application of relevant facts and law. Although vagaries
in courts decision making might wash out on average over the mine run of
cases,184 to the extent prior decisions inform later results there is some reason to
suspect that vagaries will not average out and will instead tend toward a
particular form of error that undermines patent system performance.185
Moreover, even if vagaries do wash out on average, variance in results about the
average can harm social welfare by distorting incentives for parties who are risk-
prone or risk-averse.186
The sort of forum shopping suggested by Subsection II.A.3 can exacerbate
concerns with litigation vagaries by increasing the risk of biased results when
183 See Arti K. Rai, Specialized Trial Courts: Concentrating Expertise on Fact,
17 BERKELEY TECH. L.J. 877, 877-78 (2002) (arguing for a specialized patent trial court
because difficult questions of scientific fact are likely to arise more routinely in patent law
than in virtually any other field of law); cf. Thomas F. Cotter, Reining in Remedies in Patent
Litigation: Three (Increasingly Immodest) Proposals, 30 SANTA CLARA HIGH TECH. L.J. 1, 21
(2013) (expressing a belief that the right to trial by jury contributes to the high cost of patent
litigation and to overinflated damages awards).
184 Golden, supra note 49, at 580 (noting classic arguments that [courts] average
correctness suffices to provide proper incentives).

185 At least in part because court awards of reasonable royalties can inform out-of-court
royalty rates, which can inform later court awards, a number of commentators have suggested
reasons that vagaries in the assessment of patent damages might not average out over time.
See, e.g., Einer Elhauge, Do Patent Holdup and Royalty Stacking Lead to Systematically
Excessive Royalties?, 4 J. COMPETITION L. & ECON. 535, 546 (2008) (contending that, under
various circumstances, royalty rates are likely to be suboptimal); William F. Lee &
A. Douglas Melamed, Breaking the Vicious Cycle of Patent Damages, 101 CORNELL L. REV.
385, 438 (2016) (contending that [t]he current reasonable royalty doctrine systematically
overcompensates patent holders); Mark A. Lemley & Carl Shapiro, Patent Holdup and
Royalty Stacking, 85 TEX. L. REV. 2163, 2164 (2007) (contending that patent law
systematically overcompensates patentees whose inventions are only one component of a
larger product); Jonathan S. Masur, The Use and Misuse of Patent Licenses, 110 NW. U. L.
REV. 115, 144 (2015) (discussing how use of patent licenses to calculate reasonable royalty
damages likely depresses damages more than any [upward] distortion that patentees can
introduce by inflating license prices).
186 See Golden, supra note 49, at 581.
one side to patent cases, likely that of patent holders, tends to dominate the
choice of forum.187 Further, significant discrepancies in treatment at the district
level can encourage socially wasteful strategic behavior in the form of races to
the courthouse.188
In short, there are substantial causes for concern that patent litigations high
costs, duration, and susceptibility to error or bias might undermine the patent
systems aim to promote technological progress or social welfare more
generally. The next Subsection highlights how PAEs can aggravate these causes
for concern.
2. Concerns with PAEs

To a great extent, concerns with the litigation and licensing activities of PAEs
are no more than extensions of concerns about patent litigation as a whole. PAEs
can act as helpful intermediaries, identifying potential licensees and infringers
and providing a means by which patent holders unable to afford litigation can
obtain recompense for otherwise unenforceable rights.189 But PAEs can also
exploit litigation costs, legal vagaries, and hold-up potential to tax the work
of innovators or consumers while failing to funnel sufficiently counterbalancing
compensation to inventors.190 Non-PAE patent holders can abuse the system in
similar ways. But at least on average, PAEs might be more effective at exploiting
rough edges of the patent systempartly because PAEs might be more efficient
enforcement specialists and partly because, compared to PE patent holders,
187 See Kimberly A. Moore & Francesco Parisi, Rethinking Forum Shopping in
Cyberspace, 77 CHI.-KENT L. REV. 1325, 1328 (2002) ([I]f some individuals are statistically
more likely to be plaintiffs than defendants . . ., the opportunity for forum shopping may have
biased distributional effects with a potential impact on the ex ante incentives of the parties.).
188 See id. (noting that racing to the courthouse can accelerat[e] the filing process and
bring[] to trial cases that may not have matured into court claims had they been left to the
choice of the natural plaintiff); Harvey I. Saferstein & Nathan R. Hamler, Location,
Location, Location: A Proposal for Centralized Review of the Now Largely Unreviewable
Choice of Venue in Federal Litigation, 90 OR. L. REV. 1065, 1067 (2012) (Motions to change
venue . . . are often preceded by a race to the courthouse, especially in patent
litigations . . . .).
189 See EXEC. OFFICE OF THE PRESIDENT, PATENT ASSERTION AND U.S. INNOVATION 3
(2013) (discussing how [p]atent intermediaries can play a useful social role).
190 See Brian T. Yeh, CONG. RESEARCH SERV., CRS R42668, AN OVERVIEW OF THE
PATENT TROLLS DEBATE 6 (2013) (PAEs have frequently been accused of imposing a tax
on innovation . . . .). But cf. Mark A. Lemley & A. Douglas Melamed, Missing the Forest
for the Trolls, 113 COLUM. L. REV. 2117, 2125, 2151 n.148 (2013) (finding evidence
inconclusive on the extent to which trolls increase or decrease the amounts that fund
innovation).
PAEs are likely to be less vulnerable to a patent-infringement countersuit and

less bothered by a reputation for litigiousness.191
Generally speaking, there seem to be at least three basic storylines for PAE
activities that are substantially socially detrimental:
1) Hold-Up: One concern with PAEs has been that, like a mythological troll
emerging from under a bridge, PAEs can use patent rights to ambush a
technology into which a supplier or user has become lock[ed] in, thereby
extracting a ransom that has little to do with the merits of the patented
invention.192 With a large sheaf of patents in hand and no competing business
concerns to drive the timing of litigation, certain PAEs might be especially adept
at selecting and timing patent-infringement assertions to exploit such lock-in
effects. But concerns about such potential PAE behavior appears substantially
answered by the 2006 Supreme Court decision in eBay Inc. v. MercExchange,
L.L.C.193 and its aftermath.194 In the wake of eBay, district courts have tended to
deny injunctive relief for PAEs and in cases involving complex products,195 thus
defusing concerns about lock-in. Consequently, this Articles reform proposal
specifically targets the two other PAE storylines described below.196
2) Exploiting System Vagaries for an Unmerited Windfall: Related to the
classic hold-up story but somewhat distinguishable is a lottery-ticket model
under which PAEs exploit vagaries of the patent system to pursue a large and
socially unmerited payoff.197 The value of patent rights can be very difficult to
assess, and thus, even if one makes the heroic assumption that the design of
patent law is otherwise socially optimal, there is good reason to believe that
courts assessments of patent value will be erroneous in a number of cases and
perhaps by much more than a factor of ten.198 When one considers additional,
191 John M. Golden, Patent Privateers: Private Enforcements Historical Survivors,

26 HARV. J.L. & TECH. 545, 598 (2013) (discussing potential explanations for a past tradition
of relative restraint in patent rights enforcement and acquisition); see also Cotropia, Kesan
& Schwartz, supra note 133, at 650 (noting arguments that PAEs are fundamentally
different from other plaintiffs because of their common lack of liability to patent-
infringement countersuit).
192 See, e.g., Paul J. Heald, Optimal Remedies for Patent Infringement: A Transactional
Model, 45 HOUS. L. REV. 1165, 1184, 1189 (2008) (observing that situations involving high
switching costs . . . may cover many scenarios involving patent trolls); Robert P. Merges,
The Trouble with Trolls: Innovation, Rent-Seeking, and Patent Law Reform, 24 BERKELEY
TECH. L.J. 1583, 1590-91 (2009) (The patent troll strategy is to take advantage of lock-in
that occurs as a result of [sunk cost] investments.).
193 547 U.S. 388, 394 (2006).
195 Cotter & Golden, supra note 25.
197 Lemley & Melamed, supra note 190, at 2126.
198 See Golden, supra note 14, at 2151 (noting party positions on reasonable royalties that
differed by factors of about one hundred twenty and two hundred); see also Apple, Inc. v.
real-world vagaries attendant to assessments of patent validity and scope, the

possibilities for unmerited windfalls multiply.199
3) Litigation-Cost Rent-Seeking and Harassment: In accordance with a third
storyline, some PAEsso-called bottom-feeder trolls200exploit costs of
litigation, legal advice, and uncertainty itself to extort amounts of money that
have little to do with a patent suits merits. A PAE could target a company at a
particularly vulnerable moment: conventional wisdom holds that companies
are often sued for patent infringement shortly before their initial public
offering.201 Under an alternative approach, PAEs can sue or send demand
letters to hundreds and even thousands of potential targets, including relatively
small firms and startups that are only end users of technology, such as restaurants
and hotels providing wireless Internet to customers.202 A PAE can induce
payments by such parties by offering a licensing fee that is small compared to
expected litigation costs or even the several thousand dollars commonly
necessary for a simple attorney opinion on the merits of a patent-infringement
allegation.203 In a variant of this nuisance-suit storyline, a PAE brings suit
primarily to harass its target, perhaps on behalf of a PAE sponsor. In one
suspected case of commissioned patent assertion, a company began suing
Motorola, Inc., No. 1:11-cv-08540, 2012 WL 1959560, at *3-7 (N.D. Ill. May 22, 2012)
(reporting expert opinions for reasonable royalty damages differing by a factor of one hundred
forty for one patent and by a factor of three hundred fifty for another), revd in part and
vacated in part, 757 F.3d 1286 (Fed. Cir. 2014).
199 See, e.g., Timothy R. Holbrook, Equivalency and Patent Laws Possession Paradox,
23 HARV. J.L. & TECH. 1, 46 (2009) (discussing uncertainty about patent scope resulting from
the doctrine of equivalents); Jason Rantanen, The Federal Circuits New Obviousness
Jurisprudence: An Empirical Study, 16 STAN. TECH. L. REV. 709, 717 (2013) (discussing the
real challenge involved in assessing an inventions nonobviousness).
200 Lemley & Melamed, supra note 190, at 2126 (defining bottom-feeder trolls as patent
holders that rely on the high cost of patent litigation . . . to induce the parties they sue to
settle for small amounts of money rather than pay millions to their lawyers).
201 Tom Ewing, Indirect Exploitation of Intellectual Property Rights by Corporations and
Investors: IP Privateering and Modern Letters of Marque and Reprisal, 4 HASTINGS SCI. &
TECH. L.J. 1, 70 (2012) (explaining how the prospective investor itself [can] bring the
litigation as a means for lowering the investment targets valuation price).
202 See Edward Wyatt, Obama Orders Regulators to Root Out Patent Trolls, N.Y. TIMES
(June 4, 2013), http://www.nytimes.com/2013/06/05/business/president-moves-to-curb-

patent-suits.html (noting that one PAE threatened to sue 8,000 coffee shops, hotels and
retailers for patent infringement because they had set up Wi-Fi networks for their customers);
see also Paul R. Gugliuzza, Patent Litigation Reform: The Courts, Congress, and the Federal
Rules of Civil Procedure, 95 B.U. L. REV. 279, 280 (2015).
203 See AIPLA 2015 SURVEY, supra note 18, at 29 (reporting a median estimated per-patent
charge of $15,000 for an attorney opinion on patent validity and infringement).

competitors of Nokia and Sony soon after acquiring over a hundred patents and
patent applications in which those companies had ownership interests.204
These storylines provide plausible bases for worrying about the proliferation
of PAE activity, but the net social costs of PAE activity remain subject to heated
debate.205 This Article does not attempt to resolve this debate. For this Articles
purposes, it seems enough to note two generally accepted facts. First, PAE suits
and the number of PAE-suit targets have grown substantially over the last two
decades, with annual numbers of new PAE suits and PAE-suit defendants now
tending to be in the thousands.206 Second, concerns about PAE activities
highlight systemic weaknesses that would be worrisome even without PAEs,207
including (1) limited USPTO review that leaves substantial doubt about the
validity and scope of many patents;208 (2) high litigation costs that impede access
to the courts and can foster nuisance suits and settlements;209 and (3) vagaries in
litigation outcomes that, from a social standpoint, can (a) excessively deter risk-
averse parties from suit or defense, (b) frustrate desirable settlement efforts, and
(c) encourage patent-holder rent seeking in a litigation lottery.
204 Ewing, supra note 201, at 63.

205 In March 2015, members of Congress received two letters, each signed by dozens of
scholars. The first pointed substantially to PAE activity in support of its assertion that a large
and increasing body of evidence indicates that the net effect of patent litigation is to raise the
cost of innovation and inhibit technological progress. Letter from 51 Legal and Economic
Scholars Who Study Innovation, to Congress (Mar. 2, 2015), http://www.utdallas.edu/
~ugg041000/IPScholarsLettertoCongress_March_2_2015.pdf [https://perma.cc/GAW9-
6BZQ]. The second questioned many claims of the first letter and warned that tinkering with
the engine of innovationthe U.S. patent systemon the basis of flawed and incomplete
evidence threatens to impede this countrys economic growth. Letter from 40 Economists
and Law Professors Who Conduct Research in Patent Law and Policy, to Chuck Grassley et
al., Chairman, Comm. on the Judiciary (Mar. 10, 2015), https://sls.gmu.edu/cpip/wp-
content/uploads/sites/31/2015/03/Economists-Law-Profs-Letter-re-Patent-Reform.pdf
[https://perma.cc/HL85-BW3U].
206 See, e.g., James Bessen & Michael J. Meurer, The Direct Costs from NPE Disputes, 99
CORNELL L. REV. 387, 390-91 (2014) (reporting that, over the last few years, NPE litigation
has reached a wholly unprecedented scale and scope); Cotropia, Kesan & Schwartz, supra
note 133, at 649 (observing that, [i]n the last decade, the landscape of patent litigation has
radically shifted toward enforcement by PAEs).
207 Cf. David S. Olson, On NPEs, Holdups, and Underlying Faults in the Patent System,
99 CORNELL L. REV. ONLINE 140, 150 (2014) ([I]n some ways, patent assertions by NPEs do
not raise unique problems . . . so much as they increase the severity of pre-existing
problems.).
209 See supra text accompanying notes 176-78. AIPLA survey data indicates that the costs
of defending patent-infringement suits brought by PAEs are generally roughly comparable to,
albeit often somewhat less than, those of defending against suits by other forms of patentees.
See AIPLA 2015 SURVEY, supra note 18, at 37-38.
Without need for reference to PAEs, these weaknesses of the patent system
by themselves justify this Articles proposal for patent litigation administrative
review. After all, as early as 1813, Thomas Jefferson proposed that, to better
guard our citizens against harassment by lawsuits, questions about the validity
of patents might best be turned over to a board of Academical professors
instead of the courts.210 The later institution of pre-issuance examination by
professional patent examiners was a step toward realization of Jeffersons vision
but a far from complete one. This Articles proposal can be understood as taking
a further step toward that vision.
III. PROPOSAL FOR ADMINISTRATIVE REVIEW

This Part of the Article motivates and describes a proposed framework for
early-stage administrative review of patent suits by an expert body that we call
the Patent Litigation Review Board (PLRB). We begin our discussion by
outlining the proposed PLRB process. Upon filing, district court cases would be
automatically stayed pending preliminary review by the PLRB. Parties to the
case could present arguments to the PLRB based on information available to
them prior to discovery. These arguments could involve questions of claim
construction, infringement, and enforceability as well as questions of validity.
In addressing such questions, the PLRB would take an approach to patent claim
construction identical to that of the district courts. Under a variant of a clear-
and-convincing evidence standard, the PLRB would determine whether, with
respect to any of the issues raised for preliminary review, a party sufficiently
proved its case. The PLRBs ruling would not be binding on the courts but would
be admissible in court and, we believe, likely influential. Further details
regarding the framework for early-stage administrative review appear in
Section III.B.
Section III.A makes the case for developing the PLRB framework by
discussing how PLRB review can promote improved dispute resolution and help
screen out weak claims and arguments. Section III.A uses general economic
arguments and two economic models to support the contention that PLRB
review would likely have these effects. First, a high-level model shows that even
an only roughly accurate preliminary review process can systematically increase
the expected value of higher-quality claims and decrease the expected value of
lower-quality claims. Second, a more detailed model for patent assertion and
defense demonstrates that these high-level results hold even when one accounts
for expected real-world costs of PLRB review. The two models thereby show
that PLRB review should improve patent system performance by discouraging
weaker suits and encouraging stronger suits. Further, the models are
conservative in the sense that they show PLRB review has this positive effect
210 Letter from Thomas Jefferson to Isaac McPherson (Aug. 13, 1813), in 6 THOMAS
JEFFERSON, THE PAPERS OF THOMAS JEFFERSON: RETIREMENT SERIES 384 (J. Jefferson Looney
ed., 2009).
even when one assumes parties have symmetric information and beliefs. To the
extent asymmetric information is a cause of problematic patent litigation or
patent-assertion behavior, these models most likely underestimate the benefits
of PLRB review because the models do not reflect the fact that, by inducing
valuable information exchanges and providing early administrative feedback on
party arguments, PLRB review can significantly reduce information
asymmetries before parties undertake the costly process of litigation-related
discovery.211
Sections III.B and III.C close the case for PLRB review. In addition to laying
out a specific legal framework for PLRB review, Section III.B presents a third
economic model to respond to concerns that private parties could dissipate the
social value of PLRB review by flooding the PLRB with claims and arguments.
This model shows that parties own interests in limiting private costs will likely
cause them to be selective in assertions before the PLRB, thereby helping to
preserve the PLRBs capacity to have a net positive effect on social welfare.
Section III.C follows by explaining how other recently implemented or proposed
adjustments to the patent system do not undermine the case for early-stage
administrative review.
A. Economic Analysis of Administrative Review
1. Basic Economic Theory

As discussed in Part II, key problems with U.S. patent litigation are high costs,
delay, and uncertainty. In this environment, many patent holders likely
experience undue difficulty in vindicating valid claims of infringement. At the
same time, many innovators and technology users likely experience excessive
difficulty in clearing others patents and defeating unjustified charges of
infringement.
The informational and cost advantages of a patent system with early-stage
administrative review can cause it to function better, generally speaking, for both
meritorious patent enforcers and technology users. If PLRB review meets at least
relatively minimal standards of substantive accuracy, the alteration of private-
party incentives that it effects can substantially reduce socially harmful effects
of high litigation costs, delay, and uncertainty. PLRB review can facilitate
parties early exchange of relatively high diagnosticity/cost information.212
Such review can also provide parties with a relatively quick, cheap, and
211 See infra text accompanying notes 223-29 (describing the process of PLRB review).
212 See Louis Kaplow, Multistage Adjudication, 126 HARV. L. REV. 1179, 1225 (2013)
(suggesting that, [a]s an initial, rough cut at the problem [of ordering steps in multistage
adjudication], it seems that the step with a higher diagnosticity/cost ratio should be earlier);
cf. Reilly, supra note 7, at 239 (advocating staged litigation in which [d]iscovery is
limited . . . until the plaintiff demonstrates a meritorious case by prevailing on the initial
issue(s)).
informative preview of potential results of court adjudication. The result can

include earlier convergence of parties assessments of a suits economic
potential; earlier identification of weak suits or party positions by an impartial,
third-party adjudicator; and thus earlier termination of a suit through voluntary
dismissal or settlement.213 Further, beyond simply previewing what a court
might do, the expert judgments and opinions of the PLRB can alter the balance
of post-review incentives by informing later judicial decisionmaking, helping
both to limit vagaries that might result from district court inexpertise and to
streamline later court proceedings by encouraging parties to drop at least a subset
of arguments or claims.
Of course, there can be concerns that the social costs of PLRB review will
swamp its social benefits. But even with the imperfect alignment of private and
social interests, the fact that parties must pay for that review can cause them to
restrain their use of administrative review so that, generally speaking, its social
benefits outweigh its costs. Substantial PLRB control over its procedures can
enable the agency to tailor proceedings to ensure sufficient private and social
alignment, and Subsection III.C.4 shows how sufficient alignment can be
expected under reasonable assumptions.
2. A First Economic Model for Administrative Review

A high-level economic model highlights the potentially large benefits of
preliminary administrative review even when parties are assumed to have
symmetric information but their information is incomplete in the sense that they
are uncertain about how a third-party adjudicator will rule. For such parties,
PLRB review has the value of sorting out at least some stronger and weaker legal
positions by generally increasing the party-perceived likelihood that some
stronger positions will prevail and that some weaker positions will lose.
Moreover, this high-level model suggests that benefits from PLRB review are
likely to be substantially robust against the natural fallibility of human decision-
making institutions. PLRB decisionmaking will generally increase the expected
value of stronger legal claims relative to that for weaker legal claims.
The model operates as follows:
1) A plaintiff P considers suing a defendant D.
2) The true underlying suit quality is represented by , a quantity having a

value between 0 and 1 that is known to P and D but that is not directly
213 See Robert G. Bone, Modeling Frivolous Suits, 145 U. PA. L. REV. 519, 593 (1997)
(suggesting a process of early judicial screening of frivolous suits based on a preliminary

review of the merits); Bert I. Huang, Trial by Preview, 113 COLUM. L. REV. 1323, 1335
(2013) (discussing the familiar economic model, in which an increase in information sharing
leads to convergence in the parties forecasts about the outcomeand hence to a greater
chance of settlement).
verifiable by a court. The probability that the court finds for P is (),
and this probability is increasing in . If the court finds for P, it awards
damages of > 0 to P; otherwise, it enforces a penalty 0 on P that
is transferred to D by a court award of attorney fees. P faces an
administrative review cost and incurs an additional cost for
litigating through a full court decision. D must pay to see the suit
through a decision by the court.
3) If P files suit, then D expects to receive
() + (1 ()) = ()( + ) ( )
if it pursues a court decision. We assume that D settles for otherwise.

Hence, D pursues a decision if and only if
()( + ) + ( ).
4) P will choose the settlement amount
= ()( + ) + ( ),
so that D is indifferent between settling and going to court. Assuming for

simplicity that a party chooses against continued litigation when it is a
matter of such economic indifference, we see that P will bring suit if and
only if
0 < ()( + ) + ( ) = ()( + ) ( + ).
Note that if the cost to D, , is higher than + , then P will always

bring suit.
Suppose that, after suit is announced (i.e., after the payment of , but before
settlement negotiations), PLRB review yields a signal that is informative about
the probability that the court will find for P. Specifically, assume that the
probability density of for a given value of , f ( | ), has the monotone
likelihood ratio property in . The quantity (,) is the probability of a win for
P as a function of and , increasing in both arguments. We assume that (,)
and () are equally responsive to , in the sense that their derivatives with
respect to are equal: (,) = ().214 D then expects to receive
214 This equality of derivatives with respect to generally corresponds to (,) equaling
() + g(), where () and g() fall within a range of values such that 0 () + g() 1 for
all pertinent values of and .
E[(, ) | ] + (1 E[(,) | ]) =
E [(,) | ]( + ) ( )
if it pursues a court decision. Consequently, P brings suit if and only if
0 < E[(,) | ]( + ) ( + ).
Under the above model, between a situation with early-stage administrative

review and a situation without it, the difference in expected return for P is
(E[(,) | ] ())( + ) = E[(,) () | ]( + ) (1)
= ( + ) [(,) ()] f (|)d (2)
where we have used the fact that f (|)d = 1.

Now, we note that, as (,) = (), for fixed we must have (,) ()
= (,) () for all , . Suppose that > . It follows that
[(,) ()]f ( | ) d = [(,) ()] f ( | ) d (3)
[(,) ()] f ( | ) d, (4)
where the inequality (4) follows from first-order stochastic dominance (a

consequence of the monotone likelihood ratio property) and the fact that the
quantity [(,) ()] is increasing in .
Thus, we see that PLRB review generates a difference in expected return for
P that is increasing in . As long as administrative review increases the
likelihood of P winning when its case is absolutely ironclad ( = 1) and reduces
that likelihood when Ps case is truly meritless ( = 0), it follows that there is
some case-quality value * such that all Ps with cases stronger than * return
more (in expectation) after the addition of the review stage, while all Ps with
cases weaker than * do worse. Given the fixed costs of filing, , this means
that Ps with cases of quality higher than * are more likely to file (in
equilibrium) given the review, and Ps with cases of quality lower than * are
less likely to file. Moreover, even for suits that are brought under both regimes,
the higher-quality suits return more in expectation in the presence of PLRB
review, and the lower-quality suits return less.
The results are strengthened if PLRB review reduces court costs for the side
that prevails in review, as this further increases the gains that the prevailing side
receives upon pursuing suit. The results are qualitatively unchanged if settlement
is not possible as, in that case, the comparison across litigation regimes also
hinges on the sign of equation (1) above.
3. A Second, Calibrated Model for Administrative Review

We now use a discrete-time patent-assertion model to assess the potential
costs and benefits of PLRB review in greater detail. Within the model, parties
are assumed to be rational profit maximizers subject to the limitations on options
for litigation and settlement that the model imposes in order to make
investigation manageable. In the illustrative examples featured here, parameter
values for use in the model are calibrated to be plausible but conservative real-
world figures corresponding to values suggested by pre-existing data as well as
reasonably moderate estimates of the degree to which PLRB decisions will shift
subsequent district court results. By a conservative figure, we mean one that
likely errs on the side on making the case for PLRB review more difficult. For
example, the choice of relatively low values for the stakes in our examples
presents a greater challenge for our policy proposal because it makes the parties
litigation costs for PLRB review loom comparatively larger relative to the value
of any PLRB-provided clarification of the odds of winning or losing those
stakes. By using such conservative figures in our illustrative examples, we hope
to strengthen their overall persuasive force.
a. Model Structure
Consider a discrete-time, three-period setup where patent assertion and
litigation decisions occur as indicated in Figure 1 and Table 1.
Figure 1. Discrete-Time Assertion Model with Parameters

Table 1. Model Parameters

Symbol Explanation
cost to plaintiff of pre-filing preparation
PLRB administrative fee plus plaintiff contestation costs
PLRB defendant contestation costs
district court plaintiff litigation costs
district court defendant litigation costs
probability as function of patent claim quality q that PLRB
rules substantially in plaintiffs favor overall
probability as function of patent claim quality q that PLRB
rules substantially in defendants favor overall
1 probability as function of patent claim quality q that PLRB
does not rule substantially in either partys favor overall
probability as function of patent claim
quality q that court rules substantially in plaintiffs
favor overall after PLRB does the same
probability as function of patent claim quality q that court
rules substantially in plaintiffs favor overall after PLRB
does not rule substantially in either partys favor overall
probability as function of patent claim quality q that court
rules substantially in plaintiffs favor overall after PLRB
rules substantially in defendants favor overall
transfer from defendant to plaintiff
if plaintiff substantially wins in district court
transfer from plaintiff to defendant
if defendant substantially wins in district court
s share of an expected litigation value V
that plaintiff receives in a settlement
The choice of assertion occurs at time T0. If assertion is chosen by the patent
holder P after incurring a cost , the alleged infringer D then decides at the
immediately following time T whether to settle with P or to fight the
infringement allegation. If D settles, it will pay out a quantity set by the model,
which can be conceived as equaling the product sV, where V is an expected
value associated with the litigation (e.g., the expected value to P of any payment
from D to P if the case proceeds) and s is the fraction of that value that D pays
to P in a settlement. If, instead of settling at T, D fights the infringement
allegation, the PLRB will review the case. This will cost D the amount ,
which, assuming this proceeding is decided on papers only, is the cost to D of
collecting and providing the limited supporting documentation, including
appropriate briefing, to make its arguments. The PLRB process will cost P the
amount , which includes not only the cost of collecting and providing the
limited supporting documentation to make its case for infringement to the PLRB,
but also an additional fixed fee to at least partially cover administrative

expenses.
The PLRB will not provide an extensive preliminary judgment215 in every
case. Instead, in response to party filings, the PLRB will identify and flag
particularly strong or weak positions on either side, delivering one of the
following forms of judgment:
1) Bad: Judgment is predominantly and substantially adverse to the patent

holdere.g., a judgment that one or more assertions made by the patent
holder are clearly incorrect (meritless), without substantially
countervailing conclusions in favor of the patent holder. A Bad judgment
happens with probability .
2) Good: Judgment is predominantly and substantially favorable to the

patent holdere.g., a judgment that one or more assertions made by the
patent holder are clearly correct, without substantially countervailing
conclusions in favor of the patent challenger. A Good judgment happens
with probability .
3) Inconclusive: Judgment is inconclusive in the sense that either the

PLRB draws no conclusions substantially in favor of either side or the
PLRBs conclusions are essentially balanced in giving partial and
substantially countervailing victories to each side. An Inconclusive
judgment happens with probability 1 .
At time T1 following the PLRBs decision, P decides whether to drop the suit
or to continue pursuing charges of infringement. If P chooses to continue, then
D decides how to respond at immediately subsequent time T1+. If D settles with
P, the settlement value equals the product sV1+, where s is again the settlement
fraction, this time applied to a base value V1+, which is an expected value
associated with continuing the litigation after the PLRBs decision. If, instead of
settling at T1+, D fights the infringement allegation in court, D will incur
litigation costs , and P will incur litigation costs .
Ds net expected proceeds from continuing litigation at time T1+ will depend
on both the information revealed by the PLRBs decision and the underlying
claim quality q. For example, if the PLRB issues a plaintiff-disfavoring
judgment of Bad, Ds expected court award (or loss) will equal 1
. Subtracting her cost of fighting in court yields net
expected proceeds of 1 . The
corresponding expected payoff for P of going to court after the PLRB judgment
215 Cf. Geoffrey P. Miller, Preliminary Judgments, 2010 U. ILL. L. REV. 165, 167 (using
the term preliminary judgment somewhat differently to refer to a tentative assessment . . .
based on the same sorts of information that the courts already consider on motions for
summary judgment).
of Bad is 1 . The expected payoffs in

cases of Good and Inconclusive PLRB judgments are constructed similarly
using the probabilities and associated with those states.
b. Illustrative Results
In this Subsection, we take the model and examine its implications by
applying it to eight illustrative calibrations, which are scenarios specified by
assigned sets of parameter values. An Online Appendix prepared as a
supplement to this Article presents additional examples that further indicate the
robustness of this Articles conclusions.216 Interested readers can explore the
dynamics of the model and use it in follow-on work. The Online Appendix also
includes the Mathematica code used to run calibrations.217
Generally in our examples, P will choose not to file at time T0 or not to
continue the patent-infringement suit at time T1 if, without the possibility of
settlement at immediately subsequent time T or T1+, respectively, the net
present value of suit for P is negative. This aspect of the modeled examples
corresponds to assuming that, at both of the discrete settlement times T and T1+,
D will not offer anything above $0 to settle the case unless P has a credible threat
of continuing the case absent a nonzero settlement offer.
For purposes of the examples allowing settlement, we assume that parties
settle for the expected value of the end payout in the case (e.g., the expected
value of a court award if the parties are at settlement time T1+). In other words,
for the settlement-based examples, we assume a settlement fraction s of 1 (that
is, 100%), and we apply this settlement fraction to values for V and V1+ that
equal the expected value of the amount, excluding next-stage plaintiff process
costs or , that P expects to gain if litigation continues beyond the current
settlement stage.218 Under this approach, settlement saves each party its expected
costs of undertaking the next stage of administrative or court process. Thus, in
situations where P has a credible threat of continuing litigation even without
immediate settlement, settlement is rationally desired by both parties and always
occurs at time T by taking into account the probabilistically weighted
settlements that would otherwise occur at time T1+.
To demonstrate the expected effects of PLRB review, we separately model
circumstances in which settlement is allowed and, under the model, always
occurs and in which settlement is not allowed but plaintiffs can still terminate
216 Lauren H. Cohen et al.,Troll Check? A Proposal for Administrative Review of Patent
Litigation: Online Appendix (Sept. 18, 2017), available at https://papers.ssrn.com/sol3/

papers.cfm?abstract_id=3039106 [https://perma.cc/R5SB-PEYA] (featuring additional
illustrative calibrations under the discrete time patent assertion model).
217 Id.
218 Thus, in a situation in which the PLRB has issued a plaintiff-disfavoring judgment of
Bad, the proffered settlement amount sV1+ will equal the expected court award to the P after
such a PLRB judgment: 1 .
negative net present value suits at times T0 and T1. Within these two basic sets
of circumstances, we analyze four subclasses of scenarios involving:
I. Weak Suitswithout PLRB
II. Weak Suitswith PLRB
III. Strong Suitswithout PLRB
IV. Strong Suitswith PLRB
Eight tables for the calibrated scenarios appear as Tables 2 through 9 below.
Each set of calibration results ends by reporting a value E(Ex Ante Plaintiff
Payoff) that equals the ex ante expected value for P of pursuing the relevant
claim of patent infringement against D. Because P will drop the suit at T0 if the
net present value (NPV) of filing the suit is less than zero and because P will
have incurred costs of no more than before arriving at time T0, the minimum
value of E(Ex Ante Plaintiff Payoff) in our modeled results will be . Further
discussion of the calibrations and their implications follows.
Table 2. Calibration IS: Weak Suits without the PLRB

Model Parameter Calibration Value
20,000
0
0
200,000
400,000
0%
0%
1 100%
0%
15%
0%
1,500,000
0.05 * ( + )
Settlement Fraction 100%
E(Ex Ante Plaintiff Payoff) = $188,000
Table 3. Calibration IIS: Weak Suits with the PLRB

20,000
90,000
60,000
180,000
380,000
1%
60%
1 39%
50%
15%
5%
1,500,000
0.05 * ( + )
Table 4. Calibration IIIS: Strong Suits without the PLRB

20,000
0
0
800,000
800,000
0%
0%
1 100%
0%
75%
0%
1,500,000
0.05 * ( + )
E(Ex Ante Plaintiff Payoff) = $1,095,000
Table 5. Calibration IVS: Strong Suits with the PLRB

20,000
90,000
60,000
780,000
780,000
50%
1%
1 49%
90%
75%
25%
1,500,000
0.05 * ( + )
E(Ex Ante Plaintiff Payoff) = $1,199,000
Table 6. Calibration INS: Weak Suits without the PLRB

20,000
0
0
200,000
400,000
0%
0%
1 100%
0%
15%
0%
1,500,000
0.05 * ( + )
Table 7. Calibration IINS: Weak Suits with the PLRB

20,000
90,000
60,000
180,000
380,000
1%
60%
1 39%
50%
15%
5%
1,500,000
0.05 * ( + )
Table 8. Calibration IIINS: Strong Suits without the PLRB

20,000
0
0
800,000
800,000
0%
0%
1 100%
0%
75%
0%
1,500,000
0.05 * ( + )
Table 9. Calibration IVNS: Strong Suits with the PLRB

20,000
90,000
60,000
780,000
780,000
50%
1%
1 49%
90%
75%
25%
1,500,000
0.05 * ( + )
c. Analysis of Results
In this Subsection, we analyze the calibration examples of Tables 2 through
9. These examples show the effects of PLRB review on Weak and Strong
suits. The Weak Suits are comparatively low-merit cases modeled as having a
plaintiff-win probability of 15% in the absence of the PLRB.219 The Strong
Suits are comparatively high-merit cases modeled as having a plaintiff-win
probability of 75% in the absence of the PLRB. For purposes of assigning
illustrative parameter values, P is assumed to be a patent holder and D is
assumed to be an accused infringer. Likewise, for simplicity, D does not threaten
a countersuit, whether for patent infringement, an antitrust violation, or
otherwise.
Both Weak and Strong Suits are assumed for purposes of the examples to
involve relatively low stakes by patent litigation standardsnamely, a
plaintiffs claim for an overall monetary award of $1.5 million.220 We view this
219 The Weak Suits combination of low plaintiff-win probability and relatively low stakes
compared to typical litigation costs might earn them designation as nuisance suits. See
supra notes 131, 178 and accompanying text.
220 According to a litigation study by PricewaterhouseCoopers, median court-awarded
damages in patent suits ranged from $2.0 million to $17.0 million for the twenty-year period
from 1997 through 2016, and the median damages award for the five-year period from 2012
through 2016 was $5.8 million. PRICEWATERHOUSECOOPERS, 2017 PATENT LITIGATION
STUDY: CHANGE ON THE HORIZON? 9 (May 2017), https://www.pwc.com/us/en/forensic-
conservative estimate of stakes as an assumption against interest. The lower

stakes make the provision of a net economic benefit to patent holders with
Strong Suits more difficult for a combination of reasons: (1) under the model,
the value added to a Strong Suit by PLRB review generally becomes smaller
with smaller stakes; and (2) for there to be a net economic benefit to a patent
holder from PLRB review, the value added by that review must be greater than
the additional expenses for the patent holder that PLRB review generates. The
$1.5 million value for Ps claim, like other numbers in the calibration, can be
varied using code such as that in the Online Appendix.
i. PLRB Screening of Cases with Settlement

Calibrations IS and IIS of Tables 2 and 3 model Weak Suits of relatively low
value in situations where settlement always occurs. As the expected value
E(Ex Ante Plaintiff Payoff) in Calibration IS indicates, without the PLRB the
illustrative low-merit suit has positive NPV for P of $188,000 despite only a
15% chance of an ultimate court judgment for the plaintiff. D is expected to
choose to settle the case for $208,000 in order to avoid a costly and presumably
lengthy court battle. In sharp contrast, when we introduce the PLRB in
Calibration IIS, the existence and informational contributions of the PLRB help
screen the Weak Suit, discouraging its development by P by giving the case a
negative ex ante NPV of $20,000 despite Ds propensity to settle cases to avoid
litigation. This negative NPV illustrates one of the important benefits the PLRB
would bring to the patent litigation landscapenamely, reduction in the value
of at least some low-merit suits to a point where a rational potential claimant
will not pursue them and is unable to make a credible threat of doing so.
Next, we turn to Strong Suits, relatively high-merit assertions of patent
infringement illustrated by Calibrations IIIS and IVS of Tables 4 and 5,
respectively. As might be expected in scenarios in which the plaintiff has a 75%
chance of winning a court judgment of $1.5 million without PLRB review, Ps
expected payoffs both with and without the PLRB are positive. Calibration IIIS
shows that, without the PLRB, the Strong Suit has a positive NPV for P of
$1,095,000. Again, this figure reflects the amount for which D is expected to
choose to settle the case at time T. Calibration IVS calculates the expected
payoff when the same suit is subject to PLRB review. Unlike the case of the
Weak Suit, the NPV of the Strong Suit is still positive and in fact higher when
subject to PLRB review. Indeed, at $1,199,000, the NPV for the Strong Suit is
nearly 10% larger than the NPV without the PLRB.
services/publications/assets/2017-patent-litigation-study.pdf [https://perma.cc/A9VM-
TR28]. Further, the median damages award for NPEs was nearly $11.5 million for the twenty-
year period from 1997 through 2016 and $15.7 million for the five-year period from 2012
through 2016. Id. at 16. Although the characteristic values of patent suits in general might
differ significantly from that for the subset of patent suits that result in court-awarded
damages, these figures provide at least some basis for viewing the assumption of stakes of
$1.5 million as relatively conservative.
The increase in NPV for the Strong Suit when subject to PLRB review
highlights the second main benefit the PLRB can bring to the patent litigation
landscape. In addition to discouraging a substantial number of low-merit suits,
PLRB review should encourage patent-rights enforcement and compliance in a
substantial number of cases in which patent holders have high-merit claims. This
follows from the expectation that a positive ruling from the PLRB will increase
the expected odds of winning in later district court proceedings, thereby
generating a rise in patent-infringement suit NPV that can more than pay for the
costs to the patent holder of PLRB proceedings.
We concede, however, that there will almost inevitably be patent holders with
meritorious claims who do not benefit so substantially from PLRB review. This
failure could result because the strengths of their claims are difficult to convey
in summary proceedings that precede discovery. The point that Calibrations IIIS
and IVS make is that, as a reform designed to make the patent system work better,
PLRB review will be balanced in the sense that it will have pro-patentee effects
in some situations and anti-patentee effects in others, with pro-patentee effects
tending to occur in high-quality lawsuits. The key question for PLRB review is
not who is making a claim but how strong or weak that claim can be shown to
be at the outset of litigation.
ii. PLRB Screening of Cases Without Settlement

The next four scenarios, Calibrations INS through IVNS of Tables 6 through 9,
respectively, use the same parameter estimates as those in the preceding
Subsection. But here we exclude the possibility of settlement. In these scenarios,
cases will end short of a final court judgment only if P decides unilaterally not
to pursue the case further. Examining these variants of the original scenarios not
only provides a robustness check, but also enables exploration of the impact of
settlement on the value of PLRB review. The results for Calibrations INS through
IVNS point to two primary conclusions. First, the possibility of settlement can
highly inflate the NPV of a patent suit relative to a situation in which settlement
is impossible. Second, the effect of PLRB review on the NPV of suit for a patent
holder follows the same basic pattern regardless of whether settlement is
allowed.
On the first point, note that Ps NPV for suit is commonly much higher in
Tables 2 through 5s settlement scenarios than in Tables 6 through 9s
corresponding no-settlement scenarios. The NPV of Table 6s Weak Suit
without settlement and without PLRB review is $12,000, whereas the NPV for
Table 2s corresponding Weak Suit with settlement and without PLRB review
is $188,000. For Tables 4 and 8s Strong Suit without PLRB review, the
comparative NPV figures are $1,095,000 with settlement and $295,000 without
settlement. For Tables 5 and 9s Strong Suit with PLRB review, the comparative
figures are $1,199,000 and $336,805, respectively. The only comparative
situations for which removal of the settlement option does not change the NPV
are those for Tables 3 and 7s Weak Suit with PLRB review. Both with and
without settlement these scenarios yield NPVs of $20,000, the minimum
possible value. In short, the possibility of settlement tends to increaseoften

greatlythe value of a suit to the plaintiff. PLRB review can curtail and
sometimes even prevent this inflationary effect of settlement.
Moreover, just as in scenarios where settlement occurs, PLRB review has the
apparently beneficial effect of tending to decrease the value of illustrative low-
merit suits and tending to increase the value of illustrative high-merit suits.
Tables 6 and 7 show that, in the absence of settlement, the NPV from bringing
the Weak Suit is roughly $8,000 lower with the PLRB than without it. In other
words, the PLRB causes a two-thirds decrease in the Weak Suits NPV. On the
other hand, Tables 8 and 9 show that introduction of PLRB review increases the
NPV of the Strong Suit by over 14%, from $295,000 to $336,805.
d. General Welfare and Policy Implications

Comparing the calibrations both with and without the settlement option shows
that the option to settle does not alter the basic nature of the expected positive
welfare impact of the PLRB. With or without the ability to settle, the PLRB has
a positive impact by both decreasing the value of low-merit claims and
increasing the value of high-merit claims. The first effect can help to effectively
screen out low-merit patent claims not only by causing assertions to be dropped
after PLRB review but also by discouraging the initial filing of a suit. The second
effect can encourage enforcement of such claims, strengthen the bargaining
position of relevant claim holders, and enhance the deterrent force of relevant
patents.
Note that the illustrative results reported here cover both situations for which
we have assumed 100% settlement and situations for which we have assumed
no settlement. Thus, we have investigated two corner solutions with respect to
settlement. As the ability to settle can reasonably be expected to lie somewhere
in the space between these two corners, we expect that the basic patterns for
PLRB effects that we report abovein particular, relative tendencies to
encourage high-merit suits and to discourage low-merit suitswill apply quite
generally across the spectrum of real-world settlements and assertions.
Moreover, the possibility that PLRB review will affect enforcement and
defense strategies suggests that the positive effects of PLRB review could reach
substantially beyond improvements in individual case results. By decreasing the
value of low-merit claims and increasing the chances of a relatively quick,
impartial, and expert signal about the weakness of such claims, PLRB review
can embolden defendants to pursue a no-settlement strategy or, at least, an
approach to settlement tilted more toward refusal than otherwise. As the
comparative numbers for scenarios with and without settlement show, tilting
approaches to low-merit claims toward no-settlement strategies can be expected
to depress the ex ante expected value of such claims even further than PLRB
review does on its own. Likewise, PLRB reviews tendency to increase the value
of high-merit claims can be expected to increase deterrence from associated
patents because this increase in value not only raises the expected cost of being
an enforcement target but also, by encouraging enforcement, increases the

likelihood of becoming such a target.
The illustrative examples of Tables 2 through 9 demonstrate why many
parties expecting to bring high-quality patent-infringement claims should
support a proposal for the PLRB, while many of those expecting to bring low-
quality claims should oppose the proposal. In the absence of the PLRB,
informational and process limitations of the patent system mean that low-quality
claim asserters benefit from their claims being at least partially pooled with
not distinguished fromhigher-quality claims. PLRB review reduces the extent
and duration of such pooling to the benefit of high-quality claim holders and the
detriment of low-quality claim holders. In this way, PLRB review might be
particularly valuable for cash-constrained parties such as startups or individuals
who might otherwise settle or fail to pursue claims even when their likelihood
of prevailing in court is large.
A caveat is that our analysis generally assumes, as Congress commonly
appears to assume, that the patent systems substantive legal standardsthe
standards that determine whether a claim is of high or low qualityare
sufficiently well designed that increased compliance with them tends to increase
social welfare. A patent skeptic might argue that a better reform would reduce
the value of patents and patent-infringement claims across the board or even
abolish them altogether. For the purposes of this Article, we are content to
assume that, regardless of whether society would be better off under such
comparatively drastic reforms, PLRB review will likely improve social welfare
by effectively shifting value from low-merit patent claims to high-merit ones.
B. Framework for Administrative Review

The preceding Section shows how PLRB review can counteract the negative
effects of cost, delay, and uncertainty in district court litigation by providing
relatively quick, cheap, and impartial guidance on a patent suits merits. A
further advantage of a centralized administrative review process is that it can
provide a check on forum shopping as well as an opportunity to gather
centralized information on the patent litigation systems performance. This
Section provides details on the nature of the proposed review process and how
it can be implemented.
1. Proceedings Before the PLRB

The proposed administrative review would be an automatic process that
would occur immediately after the filing of a patent suit in district court. This
automatic review might be viewed as a variant of the required review of a qui
tam False Claims Act complaint by the Department of Justice221 or the
221 See False Claims Act Cases: Government Intervention in Qui Tam (Whistleblower)
Suits, U.S. DEPT OF JUST., www.justice.gov/sites/default/files/usao-edpa/legacy/2011/

requirement of filing employment discrimination claims with the Equal

Employment Opportunity Commission before resorting to the courts.222 One can
imagine a number of ways of structuring preliminary administrative review, but
this Article proposes a specific framework as a starting point for discussion.
The proposed process would work as follows. Upon the filing of a patent-
infringement suit in a district court,223 the PLRB would be notified. Unless all
parties to the district court suit opted out, the PLRB would conduct a paper
hearing in which parties would be allowed to file documentary arguments and
evidence relating to questions of patent claim construction, infringement,
validity, and enforceability. The PLRBs approach to claim construction during
this hearing would follow that of the district courts. Although attorney
representation would be advised, PLRB proceedings would improve access to
justice by allowing parties to present relevant materials either pro se or through
qualified non-attorney representatives such as, in a business entitys case, their
chief officers or owners.224
Including time for party filings, the PLRB would have one hundred eighty
days from notification of suit to issue its determinations.225 With consent from
all parties, however, the PLRB would be able to extend this time for review. The
district court would automatically stay proceedings during the administrative
review process although, on a showing of good cause, the district court would
have power to lift the automatic stay in order to consider a motion for a
preliminary injunction, temporary restraining order, or dismissal under Federal
Rule of Civil Procedure 12(b)(6). Under existing law, the PLRB might trigger a
further stay of district court proceedings by causing the initiation of post-
issuance review at the USPTO.226
04/18/fcaprocess2_0.pdf [https://perma.cc/BVA9-YNB3] (last visited on Sept. 18, 2017)

(noting that a complaint under the False Claims Act is filed under seal for at least sixty days
to permit investigation by the Department of Justice).
222 See Filing a Charge of Discrimination, EQUAL EMP. OPPORTUNITY COMMISSION,
http://www.eeoc.gov/employees/charge.cfm [https://perma.cc/M8QB-9QKA] (last visited

Sept. 18, 2017) (All of the laws enforced by EEOC, except for the Equal Pay Act, require
you to file a Charge of Discrimination with us before you can file a job discrimination lawsuit
against your employer.).
223 Suits triggering administrative review would include declaratory judgment actions in
which a party seeks a ruling of non-infringement or patent invalidity or unenforceability.

224 See supra text accompanying note 42; cf. 5 U.S.C. 555(b) (2012) (A person
compelled to appear in person before an agency or representative thereof is entitled to be

accompanied, represented, and advised by counsel or, if permitted by the agency, by other
qualified representative.).
225 The 180-day time period is modeled on the roughly six-month period currently allowed
for patent owner response and USPTO decision on a request for inter partes or post-grant
review. See 35 U.S.C. 314(b), 324(c) (2012); 37 C.F.R. 42.107(b), 42.207(b) (2016).
226 Jonathan Stroud, Linda Thayer & Jeffrey C. Totten, Stay Awhile: The Evolving Law of
District Court Stays in Light of Inter Partes Review, Post-Grant Review, and Covered
For purposes of PLRB review, the parties would be expected to focus on

issues with respect to which they believe they can establish a decisive case
through already available documentary evidence, affidavits, and written
argument. There would be no provision for discovery. In this way, the hope is
that PLRB review can prioritize the presentation of what parties believe to be
crucial and already available evidence on potentially decisive issues, rather than
wait to present such evidence and associated argument after discovery.227
Within the original or extended stay period, the PLRB would register an
affirmative conclusion or lack thereof in accordance with a clear and
convincing plus standard. The PLRB would rule either that limits on time and
evidence do not permit a substantive determination in favor of either side or,
alternatively, that, on at least one pertinent question, the evidence at hand clearly
and convincingly establishes the correctness of one particular position and there
is no substantial likelihood that additional evidence will lead to a different
conclusion.228 For example, a patent holder might produce documentary
evidence of the workings of an accused infringing device that, in the absence of
any suggestion of potentially contradictory evidence from the accused infringer,
the PLRB would hold clearly and convincingly establishes that the device
infringesor at least satisfies one or more key elements of relevant patent
claims. In another case, an accused infringer might produce documentary
evidence that, in the absence of any suggestion of potentially contradictory
evidence from the patent holder, the PLRB would hold establishes clearly and
convincingly that an accused infringing device does not infringe because it does
not satisfy at least one requirement of relevant patent claims.229
With respect to many disputable issues, the PLRB would likely register a lack
of any affirmative conclusion if asked for its opinion. Consequently, to avoid
wasted effort and to maximize chances of an overall favorable PLRB judgment,
a party would have an incentive to be selective in asserting its best arguments.
The PLRB could encourage efficiency-promoting prioritization by regulating
the length of filings presented to it.
Business Method Post-Grant Review, 11 BUFF. INTELL. PROP. L.J. 226, 237-38 (2015)
(discussing statutory provisions and judicial practices with respect to stays of district court
litigation pending inter partes or post-grant review).
227 Cf. Kaplow, supra note 212, at 1227 (arguing that it often may make sense to organize
staging by type of evidence, perhaps begin[ning] with key documents or only a few central
witnesses).
228 The standard for a substantive determination is a variant of the standard for assessing
whether a patentee has a sufficient likelihood of success in a patent suit to justify a preliminary
injunction. Trebro Mfg., Inc. v. Firefly Equip., LLC, 748 F.3d 1159, 1165 (Fed. Cir. 2014)
(An accused infringer can defeat a showing of likelihood of success on the merits by
demonstrating a substantial question of validity or infringement.).
229 Lemelson v. United States, 752 F.2d 1538, 1551 (Fed. Cir. 1985) ([I]n order for a court
to find infringement, the plaintiff must show the presence of every element or its substantial
equivalent in the accused device.).
The PLRB would generally publish its reasoning and determinations,230 and
the PLRBs opinions and judgments would be admissible in court. On the other
hand, to avoid constitutional concerns with jury rights and the prerogatives of
Article III courts,231 the PLRBs determinations on substantive matters would
only be advisory. There would be no judicial review of the PLRBs
determination independent of continuation of the original district court action or
analogous proceeding. The trial court would conduct a trial de novo, but the
courts would give the PLRBs determinations weight in the manner prescribed
by Skidmore v. Swift & Co.232 for agency statutory interpretations. In other
words, the trial court would give those determinations weight in accordance with
the PLRBs expertise, its care in deliberation, the unanimity or consistency of
its judgments, and the quality of evidence before it.233 The provisions for trial
de novo and restriction of judicial review otherwise would be statutorily
specified, thus overriding any otherwise applicable default provisions on judicial
review under the Administrative Procedure Act.234
PLRB determinations would have effects beyond their ability to influence
later court decisions. First, if the PLRB found that, under the clear and
convincing plus standard, a patent claim is invalid based on grounds that are a
permissible basis for USPTO post-issuance review, the PLRBs determination
would give the successful challenger a right to post-issuance review, as opposed
to the mere opportunity to petition for such review. Even if the challenger does
not exercise this right, the USPTO would automatically receive notice of the
PLRBs determination and could decide sua sponte to launch an ex parte
reexamination.235 Second, if a party uses court proceedings to challenge an
adverse PLRB determination and if the party loses that challenge on grounds
identical to those invoked by the PLRB, that party would presumptively have to
pay the opposing sides associated additional court costs and reasonable attorney
230 As with court opinions, confidential information might be redacted from publicly
available versions of PLRB opinions.
231 RICHARD J. PIERCE, JR., SIDNEY A. SHAPIRO & PAUL R. VERKUIL, ADMINISTRATIVE LAW
AND PROCESS 3.7 (5th ed. 2009) ([T]he [Supreme] Court has established some boundaries
beyond which Congress cannot go concerning Article III and the Seventh Amendment.); cf.
Oil States Energy Servs. v. Greenes Energy Grp., 85 U.S.L.W. 3575 (U.S. June 12, 2017)
(granting writ of certiorari for question about the constitutionality of an administrative process
for reviewing the validity of issued patents).
232 323 U.S. 134, 140 (1944).
233 Cf. id. (describing bases for deference to administrative opinions even when not
controlling upon the courts).

234 See 5 U.S.C. 559 (2012) (stating that the Administrative Procedure Acts chapter 7,
which includes its provisions on standards for review, does not limit or repeal additional
requirements imposed by statute or otherwise recognized by law).
235 35 U.S.C. 303-304 (2012) (setting forth the Directors power to initiate
reexamination of a patent [o]n his own initiative, and any time).

fees. Third, and relatedly, if a party loses before the PLRB, the party might need
to pay costs and, in exceptional cases, reasonable attorney fees associated with
the PLRB proceedings. Fourth, PLRB determinations could factor into later
assessments of whether district court litigation involves a violation of Federal
Rule of Civil Procedure 11 or whether a court should enhance patent-
infringement damages because of egregious infringing behavior.236 Finally, if
the PLRB rules in favor of a party that lacks the capacity to appear in court to
defend itself, the PLRB would intervene in associated district court proceedings
to make its reasoning and determination part of the record and, thus, a potential
basis for denying judgment against the non-appearing party.
How would the PLRB obtain funding to cover its administrative costs?
Generally speaking, complainants who initiate the relevant suits in district court
would pay for the PLRBs preliminary review through administrative fees
determined by rule.237 As with various USPTO fees, there would be substantially
reduced fees for small entities and micro-entities that are not stand-ins for better-
monetized entities.238 Further, by request, parties, including business entities,
might qualify for in forma pauperis treatment, under which the PLRB would
waive fees and cover relevant costs through surplus from PLRB fees collected
from others or through a patent system cost added to USPTO patent issuance
and maintenance fees. If there remains concern that the fees required to pay for
PLRB review would be too high to serve interests in access to justice, PLRB
review could be supported more generally through increases to preexisting fees
associated with the obtaining and maintenance of patent rights. If USPTO fees
for post-issuance review proceedings are a reasonable guide, administrative
costs associated with the operation of the PLRB might be expected to be in the
nature of $10,000 to $30,000 per case, or about $50 million to $150 million per
236 Halo Elecs., Inc. v. Pulse Elecs., Inc., 136 S. Ct. 1923, 1932 (2016) (Awards of
enhanced damages under the Patent Act . . . are . . . designed as a punitive or vindictive
sanction for egregious infringement behavior.).
237 If fees for inter partes and post-grant review serve as guides, standard administrative
costs might be in the nature of $10,000 to $30,000. America Invents Act (AIA) Frequently
Asked Questions, U.S. PATENT & TRADEMARK OFFICE, http://www.uspto.gov/patent/laws-
and-regulations/america-invents-act-aia/america-invents-act-aia-frequently-asked
[https://perma.cc/3Y2G-M8UZ] (last visited Sept. 18, 2017). In fact, average administrative
costs for PLRB review might tend to be lower than this range suggests. Because PLRB review
would involve nonbinding assessment according to a clear and convincing plus standard as
opposed to binding judgments according to a preponderance of evidence standard, costs for
preliminary review might be significantly lower than USPTO fees for inter partes and post-
grant review. Moreover, if neither party raised any issues for review by the PLRB, there would
be no need for PLRB review and the PLRB might then waive the complainants administrative
fees.
238 Michael D. Frakes & Melissa F. Wasserman, Does Agency Funding Affect
Decisionmaking?: An Empirical Assessment of the PTOs Granting Patterns, 66 VAND. L.

REV. 67, 120 (2013).
year if we assume approximately five thousand lawsuits per year.239

Consequently, given that the USPTO grants about three hundred thousand new
patents each year,240 an increase of about $500 in total fees associated with
patent issuance would likely suffice to cover PLRB administrative costs.241
2. Reporting Responsibilities and Sunset Provision

For the PLRB to bring immediately significant clarification to the strength of
party positions in patent cases, there must be a substantial number of patent-
infringement disputes in which parties will raise issues for which the PLRB can
deliver a judgment for one party or another. Nonetheless, even if the PLRB
makes such an affirmative determination only rarely, the PLRB could make an
important contribution to policymaking by helping to clarify the state of U.S.
patent litigation and, in particular, whether courts are awash in clearly frivolous
claims. Under current conditions, assessment of the relative quality of individual
patent-infringement suits is complicated by differences between trial fora, the
common confidentiality of settlement terms,242 and selection effects associated
with settlement, which leaves only a relatively small and likely unrepresentative
subset of disputes subject to decisive court judgments.243 The PLRB will be in a
centralized, start-of-litigation position uniquely suited for gathering information
on the full cross-section of patent-infringement suits filed in district courts each
year. Thus, at worst, experience with the PLRB should provide substantial
insight into the nature of the patent litigation landscape, insight that might help
point the way toward fact-based common ground in policy debates. With a view
toward exploiting this information-gathering potential of the PLRB, Congress
could require that the PLRB provide an annual report on the state of U.S. patent
litigation as seen from the PLRBs perspective.
Particularly given that out-of-pocket litigation costs in individual patent
disputes tend to run from several hundred thousand dollars to $10 million or
more for the parties involved,244 investing as much as $30,000 per suit in PLRB

240 CALENDAR YEAR PATENT STATISTICS, supra note 73.
241 If PLRB proceedings cost about $10,000 to $30,000 each, see supra note 237 and
accompanying text, and there are still about five thousand patent suits filed in district courts
each year, see supra note 132 and accompanying text, PLRB administrative costs would total
to about $50 million to $150 million per year, an amount that could be raised by obtaining
$170 to $500 in additional fees for each of three hundred thousand issued patents.
242 Golden, supra note 49, at 550 (observing that the terms of patent-licensing
agreements . . . are generally confidential).

243 See supra text accompanying note 86 (observing that litigation-based invalidation
rates do not provide a great indication of the underlying percentage of issued patent claims
that are invalid).
244 See AIPLA 2015 SURVEY, supra note 18, at 37 (reporting median estimated attorney
fees of $600,000 per side for patent-infringement suits with Less than $1 Million at Risk
and $5 million per side with More than $25 Million at Risk).
review appears justified by the benefits that such review promises. Direct
benefits from PLRB review, such as reduced overall litigation costs, increased
accuracy in dispute resolution, and enhanced access to justice, might themselves
justify the investment. Even if these direct benefits are not as large as expected,
the systemic benefits of at least a few years of PLRB information-gathering
could make the investment more than worthwhile.
PLRB information-gathering could have diminishing returns over time,
however, and the information gathered could itself suggest there are better ways
to improve the patent litigation landscape. Thus, Congress might be wise to
consider terminating an experiment with PLRB review after a few years. Even
if termination of the PLRB is not advisable, Congress might need to tweak
statutory provisions for PLRB review to respond to private parties success in
gaming the system as originally devised.
Consequently, this Article proposes that Congress initially adopt PLRB
review on a pilot basis. Absent additional congressional action, the PLRB pilot
would terminate after a specified period, perhaps three years, that allows a
reasonable time for both the maturation of PLRB practice and the gathering of
information to aid future policymaking. If Congress wishes to restrict the impact
and burden of PLRB review further, it might apply such review only to a subset
of patent casesfor example, a sample of randomly selected cases or a subset
of cases involving particular types of subject matter such as software or business
methods.
3. Agency Location for the PLRB

There are at least two main candidates for the PLRBs location within the
administrative state. First, the PLRB could be folded into the USPTO, either as
a new division or as a branch of an expanded Patent Trial and Appeal Board
(PTAB). The PTAB already handles appeals from examiner rejections and
post-issuance trials on patent validity.245 Thus, the PTAB already has within its
jurisdiction essentially the full range of validity questions that could face the
PLRB.246 Because literal-infringement analysis parallels the novelty analysis
used in checking patent validity,247 many infringement questions would not be a
major leap from the sorts of questions the PTAB already faces. Further, the most
prominent basis for charging patent unenforceability has tended to be inequitable
conduct in the process of obtaining the patent from the USPTO,248 an issue that
245 Benjamin J. Christoff, Comment, Blurring the Boundaries: How the Additional
Grounds for Post-Grant Review in the America Invents Act Raise Issues with Separation of
Powers and the Administrative Procedure Act, 39 U. DAYTON L. REV. 111, 126 (2013).
246 Id.
247 See Bristol-Myers Squibb Co. v. Ben Venue Labs., Inc., 246 F.3d 1368, 1378 (Fed. Cir.
2001) ([I]t is axiomatic that that which would literally infringe if later anticipates if earlier.).
248 Tun-Jen Chiang, The Upside-Down Inequitable Conduct Defense, 107 NW. U. L. REV.
1243, 1244 (2013).

USPTO judges might be particularly competent to handle. Additionally, the

USPTOs general administrative structure, including its experience in recruiting
and training administrative judges and in setting fees for associated processes,
would likely ease the startup costs for PLRB implementation. Location of the
PLRB within the USPTO might also facilitate coordination of PLRB review
with parallel USPTO proceedings and could simplify funding arrangements if
the PLRB is at least partly funded by patent application or maintenance fees.
On the other hand, there are compelling reasons to resist the upfront
convenience of folding the PLRB into the USPTO. First, there is the concern
that the USPTO already strains to perform its current missions.249 The generally
increasing inflow of new patent applications250 suggests that this strain will not
let up soon. Second, the PLRBs job would be to preview likely outcomes in
proceedings in district courts, rather than before the USPTO. A PLRB located
within the USPTO might have more difficulty acting as an accurate previewer
of court outcomes as opposed to a follower of USPTO positions that courts have
not yet embraced. Third, the USPTO has traditionally (and explicitly) viewed
patent applicants and owners as its customers.251 Despite the USPTOs now
substantial docket of adversarial proceedings,252 this fact, along with the
USPTOs reliance on application and maintenance fees for its funding,253 might
justify establishing the PLRB as an independent check on the USPTOs work.
Finally, USPTO-based administrative judges already number in the hundreds,
and, for purposes of both recruiting and oversight, there might be value in the
PLRB existing as a separate, compact body focused on issues relating to patent
litigation.
If the PLRB were such a separate body, it could follow either the model of
adjudicative agencies such as the ITC or that of formally private, non-profit
corporations such as the Public Company Accounting Oversight Board
249 See, e.g., Eric B. Chen, Conflicting Objectives: The Patent Offices Quality Review
Initiative and the Examiner Count System, 10 N.C. J.L. & TECH. ONLINE EDITION 28, 30
(2008) ([S]everal challenges continue to plague the USPTO, namely the backlog of
unexamined patent applications, concerns over examiner attrition, and the increasing volume
of continuing applications and ex parte appeals . . . . (footnotes omitted)).
250 See CALENDAR YEAR PATENT STATISTICS, supra note 73 (showing that the number of
U.S. utility patent applications grew from 117,006 in 1985 to 589,410 in 2015); cf. Golden,
supra note 75, at 463 (noting that, at least for . . . utility and design patents, the general long-
term trend has been for the number of patents issued each year to increase at an accelerating
pace.).
251 Golden, supra note 38, at 1098.
252 See PTAB STATISTICS 2016, supra note 37, at 3 (noting that the USPTO received 1894
and 1683 petitions for inter partes, covered business method, or post-grant review in fiscal
years 2015 and 2016, respectively).
253 See Frakes & Wasserman, supra note 238, at 69.
(PCAOB).254 The latter model would permit the PLRB to avoid standard
federal pay scales, thereby enhancing its ability to recruit highly qualified
professionals.255 In either case, members of the PLRB could be appointed by the
President for a statutorily set term of years, subject to removal by the President
only for cause.256
4. Defusing the Flood of Claims Concern

A major concern with the addition of any early-stage review process is that it
might add complication, expense, and delay that outweigh any of its social
benefits. In the case of the PLRB, this concern has at least two prongs. First,
there is the basic concern that PLRB review adds a new stage to patent litigation
that necessarily entails some costs. Section III.As economic models indicate
that countervailing social gains can justify such costs as long as PLRB review
leads to quicker or more accurate resolution of a sufficient number of cases. A
second concern, however, threatens to undermine this response. This second
concern is that, from a social standpoint, parties might spend excessively on
argument before the PLRB and thus dissipate any social value that the PLRB
would otherwise generate. Most troublingly here, the PLRB could conceivably
encourage more filings of weak claims if patent holders come to view the PLRB
either as a useful forum for trial balloon tests of litigation prospects or as a
source of potential lottery-style windfalls in the form of occasional mistaken
judgments in favor of otherwise weak claims. The possibility of lottery-style
windfalls from PLRB errors seems substantially limited because of both the high
clear-and-convincing-plus standard for affirmative judgments and the non-
binding nature of these judgments. At least on the face of things, cause for
concern that parties will flood the PLRB with litigation trial balloons is more
fundamental.
This Subsection grapples with the possibility of a flood of trial balloons or
other overinvestment in PLRB processes by modeling individual parties
incentives to make arguments before the PLRB. The model suggests that, in a
broad range of circumstances, the private costs of making such arguments will
254 See Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 484 (2010)
(describing the PCAOB as modeled on private self-regulatory organizations in the securities
industrysuch as the New York Stock Exchange); Mick Bordonaro, Recent Development,
The Public Accounting Oversight Board After Dodd-Frank, 30 REV. BANKING & FIN. L. 474,
475 (2011) ([T]he PCAOB is a non-profit corporation under the supervision and control of
the [Securities and Exchange Commission].).
255 See Free Enter. Fund, 561 U.S. at 484-85 (noting that the PCAOBs technically private
status enables it to recruit its members and employees from the private sector by paying
salaries far above the standard Government pay scale); Bordonaro, supra note 254, at 476
([T]he PCAOB does not follow the standard federal pay scale.).
256 Cf. Free Enter. Fund, 561 U.S. at 487 (accepting the belief that the Commissioners
cannot themselves be removed by the President except under the Humphreys Executor
standard).
cause costs of PLRB review to be self-regulating in the sense that rational

parties self-interest will motivate them to restrain investments in PLRB
processes enough to preserve the PLRBs promise as a means for improving
social welfare. Admittedly, however, the model also suggests that, under a
limited set of circumstances, private parties use of PLRB proceedings would, if
otherwise unchecked, drive the social costs of those proceedings above their
social benefits. This possibility provides reason to empower the PLRB to
regulate its proceedings in ways at least comparable to those available to district
courts or the USPTO. For example, by imposing page limits on filings,
demanding that parties make assertions with particularity, or limiting the number
of issues a party can present for review, the PLRB can do much to counter
private tendencies toward socially excessive argument.
This Subsections model for PLRB review works as follows. First, in light of
the clear-and-convincing-plus standard for PLRB review, we model two basic
ways that the PLRB can rule on an issue raised by a party petitioner: (1) the
PLRB can agree that the petitioner should prevail based on the information at
hand (an agreement hereinafter described as an affirmative ruling); or (2) the
PLRB can rule that existing evidence is inconclusive. Let us assume that the
probability of the PLRB agreeing that the petitioner should prevail is given by
the nonnegative value p. For simplicity, let us also assume that an inconclusive
ruling by the PLRB does not generate any information that adds or destroys
value for society or the individual parties.257 On the other hand, if the petitioner
obtains a favorable ruling, the petitioner will effectively win an amount equal to
Vpetr because of an increase in the expected value of the petitioners side of the
litigation. There are at least two potential channels for contributions to Vpetr:
(1) an increase in the petitioners likelihood of at least partially prevailing in any
subsequent district court litigation, and (2) an increase in the likelihood of
settlement on terms relatively favorable to the petitioner.
Of course, there is also a price for the possibility of winning in the PLRB
proceeding: the cost to the petitioner Cpetr of raising the issue in question and
then litigating it before the PLRB. Under the model, parties act as rational profit-
maximizers in the sense that they raise an issue for PLRB review only when the
probability of winning times the value of winning exceeds or equals the cost of
making and pursuing the relevant petitioni.e., only when pVpetr Cpetr or,
alternatively stated, p Cpetr/Vpetr.
The model assumes that society gains from the clarification that an
affirmative PLRB ruling brings. The value of that gain is Vsoc, and the probability
257 One might expect that an inconclusive ruling by the PLRB will have a negative effect
on the expected value of the litigation for the petitioner: the case will thereby become one in
which the petitioner has failed to prevail before the PLRB, rather than one in which the
petitioner only had a probability of failing to prevail before the PLRB. But because the
standard for prevailing before the PLRB is intended to be very demanding, we assume any
such negative effect to be negligible to a first approximation.
of that gain is p, the previously indicated probability of an affirmative ruling.

This gain will come at a cost to society Csoc, however. This cost Csoc is modeled
as including the cost to the petitioner Cpetr of engaging in PLRB proceedings, the
total cost Cresp to any other parties of responding, and any cost Cpub to the public
of conducting the proceedings that the parties do not bear.258 The net social
benefit as a result of the petitioners pursuit of an issue before the PLRB is =
pVsoc Csoc. Thus, under this analysis, society only loses from PLRB
proceedings if, for an appropriately representative petition, pVsoc < Csoc or,
alternatively stated, p < Csoc/Vsoc.
Hence, under the model, for the raising of an issue before the PLRB to be
rationally in a partys interest but contrary to societys interest, the value of p,
the probability of the petitioning partys prevailing before the PLRB, must lie
within a doubly restricted range: Cpetr/Vpetr p < Csoc/Vsoc. Under some
circumstances, Csoc/Vsoc could be less than or equal to Cpetr/Vpetr , with the result
that there are no values of p that satisfy the twin conditions. This situation can
arise when there are strong positive externalities to the PLRBs providing a
preliminary judgment on a particular pointfor example, the invalidity or
limited scope of a patent claim whose validity and breadth have significant
implications for competitors of the petitioner as well as the petitioner itself.259
On the other hand, one can also anticipate that, when a party chooses to pursue
litigation before the PLRB, the party will often be litigating an issue that, if the
party prevails, will produce disproportionate benefit, relative to the rest of
society, for that individual partyfor example, by effecting a wealth transfer
between the parties to the case without significantly benefiting society at large.
Consequently, given that the cost to society of PLRB proceedings Csoc includes
and therefore generally exceeds Cpetr, one can imagine situations where Vpetr is
so sizable relative to Vsoc that Csoc/Vsoc > Cpetr/Vpetr. Under such circumstances,
there is a range of p values that leads to PLRB review having a net negative
value for society.
But it is important to recall that, under the model, p values that can lead to net
negative social welfare effects are always capped by the value Csoc/Vsoc, the ratio
258 One could argue that payments to individuals such as attorneys, expert witnesses, and
PLRB employees are fundamentally just wealth transfers that presumptively lack a first-order
effect on overall social welfare. See Herbert Hovenkamp, Antitrusts Protected Classes, 88
MICH. L. REV. 1, 14 (1989). But for purposes here, policymakers can be viewed as primarily
seeking to assess whether PLRB review can achieve net value through cost savings relative
to district court litigation costs or other process costs that will be incurred without PLRB
review. In substantial part because such cost savings form the most readily estimable and,
likely, surest benefit of PLRB review, assessing when and whether the analogous costs of
PLRB review outweigh the expected benefit of PLRB review seems a reasonable approach
for first-cut cost-benefit analysis.
259 Golden, supra note 191, at 616 ([A] patent challenge can generate significant positive
externalities that are not positively reflected in a challengers incentives.).

of the cost to society of PLRB proceedings divided by the social value added by
those proceedings. If we assume that a PLRB ruling declaring a partys position
to be weak or strong will often lead to the associated issue dropping out of further
litigation and that the ratio between the costs of PLRB proceedings and the costs
of litigation before the district courts will roughly track the ratio of costs
associated with current USPTO inter partes review proceedings and the costs of
litigation before the district courts, then we have reason to believe that the ratio
Csoc/Vsoc will equate to a value of 10% or lower. In other words, because district
court patent litigation seems, generally speaking, to result in litigation costs that
are at least about ten times higher than those characteristic of administrative
proceedings,260 litigation cost savings alone appear likely to cause the ratio
Csoc/Vsoc to take a value that is no more than about 10%. In short, there is reason
to believe that PLRB proceedings will produce net positive value for society as
long as parties restrain themselvesor are restrained through appropriate rules
governing proceedingssuch that the overall success rate of their filings is at
least about 10%.
Such private-party discretion in raising issues in preliminary proceedings is
more than theoretically plausible. In the face of an arguably more demanding
standard for success than the PLRBs proposed clear-and-convincing-plus
standard, patentees already restrain their filing of motions for preliminary
injunctions so that such motions have a success rate of nearly 20%.261 This real-
world example suggests that, as long as PLRB policies and the PLRBs decision-
making record make clear that a petitioner has only a very limited chance of
obtaining a favorable affirmative judgment, parties are likely to restrict their use
of PLRB proceedings sufficiently to make PLRB review socially worthwhile.
The PLRB could adjust procedure to encourage or effectively force further
restraint as needed. Hence, worries about private parties overinvesting in PLRB
processes do not appear to provide a strong basis for rejecting this Articles
proposed reform.
C. Complement or Substitute for Other Reforms

A different potential objection to this Articles proposal is that other already-
adopted reforms or reform proposals render institution of PLRB review
unnecessary.262 In response, this Section discusses such alternatives and the
work that they leave for a PLRB.
260
See AIPLA 2015 SURVEY, supra note 18, at 37-38.
261
Kirti Gupta & Jay P. Kesan, Studying the Impact of eBay on Injunctive Relief in Patent
Cases 12 tbl.2 (Univ. of Ill. Coll. of Law Legal Studies Research, Paper No. 17-03, 2015)
http://ssrn.com/abstract=2629399 [https://perma.cc/4UFK-VBQP] (reporting that over 17%
(115 of 663) of preliminary injunction motions in patent cases have been granted since eBay
was decided in 2006).
262 Cf. F. Scott Kieff & Henry E. Smith, How Not to Invent a Patent Crisis, in REACTING
TO THE SPENDING SPREE: POLICY CHANGES WE CAN AFFORD 55, 55 (Terry L. Anderson &
In 2011, Congress made multiple adjustments to the patent systems

mechanisms for dispute resolution. Early in that year, Congress launched a
Patent Pilot Program that, in certain districts, cultivates judicial expertise by
preferential assignment of patent cases to a subset of district judges.263 Later that
same year, Congress passed the AIA, which, as noted earlier, both expanded
opportunities for USPTO post-issuance review of patent validity and tightened
joinder rules to limit the number of defendants that a patent holder may sue in a
single case.264
The courts have also been active in making adjustments. The Supreme Court
and Federal Circuit have issued decisions that, among other things, have
(1) tightened enforcement of the patentability requirements of subject-matter
eligibility,265 nonobviousness,266 claim definiteness,267 and adequate
disclosure;268 (2) emphasized district courts discretion to deny injunctive
relief;269 (3) tightened or re-emphasized requirements for proving patent-
infringement damages270 and, on the other side of a patent-infringement suit,
inequitable conduct before the USPTO;271 (4) increased opportunities for
Richard Sousa eds., 2009) (noting rapid, and we would argue excessive, changes that have
already occurred in the courts).
263 See Patent Cases Pilot Program, Pub. L. No. 111-349, 124 Stat. 3674, 3674 (2011);
Anderson, supra note 159, at 691.

265 See John M. Golden, Flook Says One Thing, Diehr Says Another: A Need for
Housecleaning in the Law of Patentable Subject Matter, 82 GEO. WASH. L. REV. 1765, 1768-
74 (2014).
266 See Jason Rantanen, The Federal Circuits New Obviousness Jurisprudence: An
Empirical Study, 16 STAN. TECH. L. REV. 709, 764 (2013) (concluding that appellate case law
on nonobviousness has effectively tightened the nonobviousness requirement).
267 Nautilus, Inc. v. Biosig Instruments, Inc., 134 S. Ct. 2120, 2130 (2014) (abrogating an
apparently loose standard for patent claim definiteness that had appeared in Federal Circuit
case law).
268 Ariad Pharms., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1340 (Fed. Cir. 2010) (en banc)
(We now reaffirm that [35 U.S.C.] 112, first paragraph, contains a written description
requirement separate from enablement . . . .).
269 eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 394 (2006) (emphasizing that the
decision whether to grant or deny injunctive relief rests within the equitable discretion of the
district courts).
270 See, e.g., LaserDynamics, Inc. v. Quanta Comput., Inc., 694 F.3d 51, 78 (Fed. Cir.
2012) (finding abuse of discretion in the district courts admission of certain evidence on
reasonable royalty damages); Douglas G. Smith, The Increasing Use of Challenges to Expert
Evidence Under Daubert and Rule 702 in Patent Litigation, 22 J. INTELL. PROP. L. 345, 350-
73 (2015) (discussing developments in the admissibility of expert testimony in patent cases).
271 John M. Golden, Patent Laws Falstaff: Inequitable Conduct, the Federal Circuit, and
Therasense, 7 WASH. J.L. TECH. & ARTS 353, 371-72 (2012) (discussing a shift toward a
leaner doctrine of inequitable conduct).
attorney fee shifting;272 (5) broadened application of a statutory rule of

construction to effectively require narrower interpretations of many existing
patent claims;273 (6) overruled decisions erecting hurdles to bringing or
triumphing in declaratory judgment actions challenging patent validity or
scope;274 and (7) used writs of mandamus to order transfer of patent cases to new
districts, particularly in cases in which the original district was the Eastern
District of Texas.275 Trial courts have taken additional measures. Starting in
2000 with the Northern District of California,276 various district courts and
judges have adopted local rules or standing orders specifically directed toward
managing patent litigation.277
States have also taken measures to regulate patent assertion. A majority of
states have passed laws specifically targeting patent demand letters, typically by
criminalizing patent-infringement allegations made in bad faith and by requiring
alleged violators to post a bond for potential penalties.278 Even in states without
272 Octane Fitness, LLC v. ICON Health & Fitness, Inc., 134 S. Ct. 1749, 1755 (2014)
(holding that a Federal Circuit framework for when attorney fees may be awarded was
unduly rigid).
273 Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1349 (Fed. Cir. 2015) (en banc in
relevant part) (overruling precedent characterizing as strong the presumption that a [patent
claim] limitation lacking the word means is not subject to [35 U.S.C.] 112, para. 6).
274 Medtronic, Inc. v. Mirowski Family Ventures, LLC, 134 S. Ct. 843, 846 (2014)
(reversing a Federal Circuit holding that a licensee bears the burden of proving non-
infringement in a declaratory judgment action); MedImmune, Inc. v. Genentech, Inc., 549
U.S. 118, 137 (2007) (holding that a patent licensee was not required . . . to break or
terminate its 1997 license agreement before challenging the patent in a declaratory judgment
action).
275 Paul R. Gugliuzza, The New Federal Circuit Mandamus, 45 IND. L. REV. 343, 346
(2012) ([T]he Federal Circuit has, on ten occasions since December 2008, granted
mandamus to order the U.S. District Court for the Eastern District of Texas to transfer a patent
case.).
276 Pelletier, supra note 109, at 478 ([T]he Northern District of California is the
acknowledged model and pioneer of local patent rules, having adopted its first version in
2000.).
277 MENELL ET AL., supra note 104, at 2-14 to 2-15 (reporting that patent local rules were
developed to facilitate efficient discovery and that they promote efficient case
management); La Belle, supra note 173, at 63 (Today, thirty district courts in twenty
different states have comprehensive local patent rules, and many more individual judges have
adopted local-local rules or standing orders that apply to patent cases in their courts.).
278 Ryan Davis, Patent Troll Targets Getting Boost from State Laws, LAW360 (Nov. 24,
2015, 8:13 PM), https://www.law360.com/articles/731287/patent-troll-targets-getting-boost-

from-state-laws (observing that [d]ozens of states have recently passed laws making it a
crime under state law to allege patent infringement in bad faith and that fifteen states include
a bond provision).
such laws, alleged infringers have begun invoking state consumer protection
laws in counterclaims against patentees.279
But these now relatively longstanding adjustments have proven inadequate to
answer continuing concerns that patent assertion has become a drag on
innovation. The rate of new patent-suit filings in district courts remains more
than double that of the year 2000,280 and these suits are now supplemented by a
comparable rate of new filings in USPTO post-issuance proceedings.281 Part of
the failure of past reforms reflects adjustments by patent holders. For example,
patent holders have deployed multiple responses to new joinder limitations, such
as increasing the number of filings against individual defendants who would
have previously been named in a single complaint; employing multidistrict
litigation (MDL) procedures to bind cases for pre-trial activities; and
incorporating in Delaware to pursue defendants in a forum where consolidation
of cases is likely because of the forums relatively small size.282
More recent changes to the Federal Rules of Civil Procedure offer little
assurance of relief. In 2015, the Supreme Court approved new Federal Rules of
Civil Procedure that seek to control[] the expense and time demands of
litigation and to promote prompt and efficient resolutions of disputes.283
Steps to these ends include shortening the default deadline for a trial judges
mandatory scheduling order;284 limiting authorized discovery to nonprivileged
matter proportional to the needs of the case;285 and apparently tightening
patent-suit pleading requirements by eliminating the ability of patentees to rely
on Form 18, a model complaint for patent-infringement suits.286 But these
279 Id. (noting use of state consumer protection statutes by [s]ome accused infringers).
280 See supra note 132 and accompanying text (noting that the number of patent suits filed
in district courts rose from about two thousand in the year 2000 to over four thousand five
hundred in 2016).
281 See PTAB STATISTICS 2015, supra note 90, at 3 (reporting that, from fiscal year 2014
to fiscal year 2015, the number of petitions for inter partes review, post-grant review, and
covered business method review increased from 1489 to 1897); see also PTAB STATISTICS
2016, supra note 37, at 3 (reporting that, in fiscal year 2016, 1683 petitions for inter partes
review, post-grant review, and covered business method review were filed).
282 Smith & Transier, supra note 15, at 231-32.
283 CHIEF JUSTICE OF THE SUPREME COURT OF THE UNITED STATES, 2015 YEAR-END
REPORT ON THE FEDERAL JUDICIARY 6.

284 Redline of Civil Rules Amendments 8 comm. note, U.S. COURTS (Dec. 1, 2015),
http://www.uscourts.gov/file/18905/download [https://perma.cc/65CB-F6UY] (The time to

issue the scheduling order is reduced to the earlier of 90 days (not 120 days) after any
defendant has been served, or 60 days (not 90 days) after any defendant has appeared.).
285 CHIEF JUSTICE OF THE SUPREME COURT, supra note 283, at 10.
286 Id. at 49 (showing abrogation of Rule 84); see Bultman, supra note 32 (The changes,
a rewrite of the Federal Rules of Civil Procedure . . . eliminate a rule that allowed filers of
patent suits to rely on a bare-bones model complaint.).
changes appear to leave much to trial judges discretion287 and seem unlikely to
have great effects in the many district courts that already have special local rules,
standing orders or practices to regulate pretrial processes in patent cases.288
Additional adjustments might be forthcoming from Congress or the courts. In
recent years, members of Congress have introduced various bills focused on
patent litigation reform289 or the pre-litigation demand letters290 that have
attracted the attention of state legislators.291 Litigation reform bills have
proposed changing aspects of litigation such as pleading requirements, rules for
attorney-fee shifting, the rules regulating allowable venues for district court
litigation, and the scope of allowable discovery.292 But, so far, efforts to enact
such bills have stalled.293
In contrast, in May 2017, the Supreme Court effectively enacted venue
reform by overturning longstanding Federal Circuit precedent that had adopted
a liberal understanding of the statutory requirements for venue.294 The change
287 See Alexsam, Inc. v. IDT Corp., 715 F.3d 1336, 1342 (Fed. Cir. 2013) (A district
courts decision to sanction a litigant [for a discovery violation] under Rule 37 is reviewed
for abuse of discretion.); Abbott Point of Care Inc. v. Epocal, Inc., 666 F.3d 1299, 1302 (Fed.
Cir. 2012) ([T]his court reviews the district courts denial of discovery, an issue not unique
to patent law, for abuse of discretion, applying the law of the regional circuit . . . . (quoting
Patent Rights Prot. Grp., LLC v. Video Gaming Techs., Inc., 603 F.3d 1364, 1371 (Fed. Cir.
2010))); Kaplow, supra note 212, at 1285-86 (suggesting that the plausibility pleading
standard might effectively leave much up to the idiosyncratic views of individual district
judges).
288 See supra notes 276-77 and accompanying text (discussing local rules and standing
orders adopted by district courts for patent cases); cf. Robert G. Bone, Twombly, Pleading
Rules, and the Regulation of Court Access, 94 IOWA L. REV. 873, 878 (2009) (contending that
the Courts plausibility standard marks only a modest departure from notice pleading).
289 See Gugliuzza, supra note 202, at 283 (Of the fourteen patent reform bills introduced
in the 113th Congress, five focused specifically on patent litigation.).

290 See id. at 283 n.25 (discussing bills that would mak[e] it illegal to send in bad faith
a letter threating [sic] patent infringement litigation).

291 See supra text accompanying note 278.
292 See, e.g., Protecting American Talent and Entrepreneurship Act of 2015, S. 1137, 114th
Cong.; Innovation Act, H.R. 9, 114th Cong. 2 (2015); Gugliuzza, supra note 202, at 283.
293 Sources Say Patent Bills Not Right-Sized But Goodlatte, Pro-Bill Lobbyists Keep
Fighting, 90 PAT. TRADEMARK & COPYRIGHT J. (B.N.A.) No. 2234, at 3624, 3624 (2015)
(reporting indications that patent litigation reform legislation was unlikely to move forward
in its current form).
294 See TC Heartland LLC v. Kraft Food Brands Grp. LLC, 137 S. Ct. 1514, 1517 (2017)
(concluding that a domestic corporation resides only in its State of incorporation for
purposes of the patent venue statute); Mark Curriden, The Rocket Docket Starts to Fizzle:
Patent Infringement Cases Filed in East Texas Federal Courts Down for First Time in 3
Years, HOUS. CHRON., Jan. 30, 2017, at B1, B2 (reporting that, if the Supreme Court reads
venue provisions narrowly, it would likely mean fewer cases filed in East Texas).
will likely have only limited impact on the litigation concerns motivating this
Articles proposal, however. Neither this change to the understanding of venue
requirements nor other proposed reforms obviate the value promised by PLRB
review. Indeed, those reforms would not provide for the sort of automatic and
impartial early-stage review of substantive questions in patent litigation that the
PLRB would make availableand hence, those other reforms cannot offer the
informational advantages for individual parties and society that PLRB review
provides. In short, even if various alternative reforms are implemented, this
Articles framework for patent litigation administrative review will still promise
to substantially improve how the patent system operates.
CONCLUSION
The time has come to take seriously a variant of Jeffersons proposal for a
board of Academical professors for patents295here, a proposed Patent
Litigation Review Board of experienced patent professionals with broad access
to scientific and technical expertise.296 Early-stage PLRB review of patent suits
can improve the patent systems immediate economic performance and provide
information that enables further improvements. Patent litigation in the United
States currently bears many hallmarks of a process ripe for, and indeed marked
by, opportunistic behavior. High litigation costs, long delays in obtaining
clarifying decisions, and substantial continuing limitations on USPTO review
suggest that a robust process of early-stage administrative review can mitigate
current problems with patent litigation. Such administrative review seems
particularly likely to be justified on a cost-benefit basis to the extent it focuses
on the tails of party assertions in patent litigation: discouraging the weakest
assertions and encouraging the strongest ones. This Article has shown that
multiple economic models, using realistic figures for the costs of review, support
the argument that early-stage administrative review will improve the patent
system.
On the other hand, dramatic litigation reforms such as the institution of PLRB
review can sometimes generate unexpected effects. Consequently, this Article
recommends adopting the proposed framework on a trial bias. A sunset
provision would require policymakers to reevaluate the framework within a few
years. Innovative policymaking with a continuing commitment to information
gathering and reevaluation are crucial to ensuring the optimal performance of
dynamic legal regimes like the patent system. This Articles proposal for PLRB
review is an ideal first step along this path.
295 See supra text accompanying note 208.

296 Cf. Golden, supra note 63, at 327 (arguing that, for claim construction, the optimal
perspective is likely to be that of a patent attorney, albeit one who has not only legal expertise
but also access to the technical knowledge of an artisan).

SSRN Id3069819

Uploaded by

Copyright:

Available Formats

SSRN Id3069819

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSRN Id3069819

Uploaded by

Copyright:

Available Formats

________________________________________________________________________

THE UNIVERSITY OF TEXAS

Troll Check? A Proposal for

Electronic copy available at: https://ssrn.com/abstract=3069819

LAUREN COHEN, JOHN M. GOLDEN, UMIT G. GURUN

INTRODUCTION ............................................................................................. 1777

L.E. Simmons Professor, Harvard Business School; Research Associate, National

Professor of Accounting and Finance, University of Texas at Dallas School of

Management; Research Economist, National Bureau of Economic Research.

Department of Economics, Harvard University; Affiliate, Harvard Center of Mathematical

Electronic copy available at: https://ssrn.com/abstract=3069819

B. Social Welfare Concerns with Patent Litigation ....................... 1801

The patent system is commonly justified as a way to promote social welfare

Electronic copy available at: https://ssrn.com/abstract=3069819

for patent reform).

4 The Innovation Act, HOUSE JUDICIARY COMMITTEE https://judiciary.house.gov/the-

innovation-act/ [https://perma.cc/XG8J-N69X] (last visited Sept. 18, 2017).

free! and subheading Time to fix the patent system).

courts/court-role-and-structure [https://perma.cc/Z63M-5EPU] (last visited Sept. 18, 2017).

MACHINA, supra note 17.

LAW360 (July 6, 2017, 9:37 PM), https://www.law360.com/articles/941916/gilstrap-s-venue-

HANDBOOK ON THE ECONOMICS OF INTELLECTUAL PROPERTY LAW (Peter Menell, David

HARV. J.L. & TECH. 545, 605 (2013) (collecting cases).

tightened the requirement of patentable subject matter . . . and the requirement of

30, 2015, 3:55 PM), https://www.law360.com/articles/717900/stricter-patent-pleading-

of a typical patent application).

USPTO regulation prescribing the broadest reasonable construction standard).

review would provide a systematic backup to USPTO review and a substantially

I. PRIMER ON PATENTS AND USPTO REVIEW

A. Patent Rights and Their Enforcement

271(a) (listing acts that constitute patent infringement).

[d]irect infringement is a strict-liability offense for which a defendants mental state is

knowledge of the relevant patent is required for indirect infringement).

principles of equity to prevent the violation of any right secured by patent).

matter itself, entities variously called patent aggregators, patent assertion

B. The Patent Document and USPTO Review

63 John M. Golden, Construing Patent Claims According to Their Interpretive

particular patent.65 Under the doctrine of equivalents, however, there is often

utility requirement is relatively low.).

description and enablement requirements).

requirement of patent claim definiteness).

YEARS 1963-2015 (2017) [hereinafter CALENDAR YEAR PATENT STATISTICS],

ACCOUNTABILITY REPORT 14 (2015), http://www.uspto.gov/sites/default/files/documents/

non-patentability by a preponderance of evidence.80 The USPTO might further

395, 411-13 (2009).

the validity or infringement of a patent).

requires an invalidity defense to be proved by clear and convincing evidence).

whose characteristics vary substantially from allowed patents in general).

(The risk that a patent will be declared invalid is substantial.).

through which the validity of patent claims might be challenged or clarified.88

7 (2015) [hereinafter PTAB STATISTICS 2015], http://www.uspto.gov/sites/default/files/

93 See MERGES & DUFFY, supra note 60, at 1047.

printed publications.96 Post-grant review and covered business method review

II. THE PATENT LITIGATION LANDSCAPE

96 See supra text accompanying notes 91-92.

of Patent Trolls, 34 REGULATION 26, 26 (concluding that lawsuits brought by NPEs

A. Recipe for Trouble

1. Litigation Complexity, Longevity, and Cost