This brief is also available in
PDF format.

UNITED STATES DISTRICT COURT

FOR THE SOUTHERN DISTRICT OF NEW YORK

THE AUTHORS GUILD, Associational Plaintiff, and HERBERT MITGANG, BETTY
MILES, DANIEL HOFFMAN, PAUL DICKSON, and JOSEPH GOULDEN, Individually and
on Behalf of All Others Similarly Situated,

Plaintiffs,

v.

GOOGLE, INC.,

Defendant.

Civil Action No. 05-CV-8136

INTEREST OF AMICUS CURIAE

Public Knowledge is a non-profit public interest organization devoted to
preserving the free flow of information in the digital age. Through
day-to-day policy advocacy, public education, and legal action, Public
Knowledge works for a balanced copyright regime that promotes creativity,
innovation, competition, and civic discourse, and limits corporate
control over the flow of knowledge. Such a regime should ensure that the
rights of the artist are protected while providing both artists and users
with the most open, competitive artistic marketplace possible.

Among the many copyright issues important to Public Knowledge is the fate
of orphan works—works whose copyright holders are unknown or who
cannot be found. In such a situation, the work becomes inaccessible for
reprinting, adaptation, or other future uses, since the owner is not
present to negotiate or grant a license. Despite the owner’s
absence, potential republishers, adapters, or other users of a work will
often decide to let such works lay dormant because of the potential for
extensive statutory damages should an owner appear at a later date.
Public Knowledge has been involved for years in seeking legislative
solutions to this problem, with the goal of enabling broader public
access to and use of these works while seeking to find ways for users of
orphan works to find their owners.

Google Book Search is an invaluable tool for making knowledge and
creative works available to the general public. This service makes it
possible for the researchers, readers, and anyone else who is interested
to find books on nearly any topic, many of which would have lain dormant
in the back of a university library or been otherwise undiscovered by all
but the most diligent and resource-endowed researcher. Public Knowledge
believes that the uses Google has made in creating and offering Google
Book Search—specifically, the scanning, indexing, and displaying of
short excerpts of books—are already lawful, and require no licenses
or permission.

While public access to orphan works is a goal towards which
amicus and the parties all strive, it is necessary that this
access be open to all comers on a level playing field. Access through one
rights organization governed by non-representative authors and publishers
and a single distributor is not truly access at all, and such a solution
ultimately benefits neither artists nor the public. Further, because a
solution to the problem of orphan works will require changes to the
operation of copyright law, those changes must be made by the legislature
through a public process involving all stakeholders rather than through a
judicial process where both parties stand to gain from licensing the
rights of absentees.

The precarious balance of interests involved in debates around orphan
works also make it essential that any orphan works solution have a legal
basis and integrity that cannot be impugned. Public Knowledge therefore
submits this brief in the hope that any settlement will not prejudice a
fair and effective solution to the problem of access to orphan works.

Introduction

Copyright law seeks to strike a balance between the rights of the
creators and the rights of users of copyrighted works. See Sony Corp.
of Am. v. Universal City Studios, Inc.,464 U.S. 417, 431-32 (1984)
(quoting Twentieth Century Music Corp. v. Aiken, 422
U.S. 151, 156 (1975)). The current balance struck by the law likely does
not permit commercial entities to digitally display the full text of
in-copyright books to the general public without the permission of the
rightsholder, even if he or she cannot be found or is unaware of his or
her rights. The proposed settlement would constitute an extraordinary
shift in this balance for both orphan and non-orphan books.^[1] Should the settlement be
approved, books that currently cannot be safely republished or digitally
distributed without first acquiring the owner’s permission would be
made available to both institutions and individual users.

While such a result may benefit the public interest by increasing
available access to knowledge, this access will be provided only through
a single bottleneck comprising the proposed Book Rights Registry
(“BRR”) and Google. In essence, the default rules of
copyright will be rewritten without the input or consent of Congress, and
only for Google and the BRR. The law does not support such an outcome:
the mechanism by which the parties currently hope to license the ability
to display digitized books exceeds the limits of class action procedure
and produces anticompetitive limitations on access to orphan works.

The proposed agreement risks stretching class action procedure too far in
its attempt to achieve in a settlement what is impossible in private
contract. The settlement’s use of class action procedure leads to
detrimental, anticompetitive limitations on the availability of orphan
works. The sought-after access to orphan works would be provided to the
public by only one party, with legal barriers erected against any others
attempting to provide access to the public on different terms. Further,
the provisional class purports to include members who are defined by
their absence and logically defy representation, and the proposed
settlement seeks to release liability for future acts not contemplated in
the activities leading up to the complaint.

Public Knowledge respectfully requests that the Court not approve the
settlement unless its flaws in antitrust effects, class representation,
and scope of release can be remedied.

I. The settlement as drafted creates antitrust and competition problems
in violation of the law and sound policy.

This Court should carefully consider the proposed settlement’s
effects on competition. In approving a settlement, a court must account
for the interests of third parties when the settlement will affect their
rights. In re Masters Mates & Pilots Pension Plan and IRAP
Litig., 957 F.2d 1020, 1025-26 (2d Cir. 1992) (“[w]here the
rights of third parties are affected, . . . their interests too must be
considered” during the court’s evaluation of “the
fairness, reasonableness and adequacy of the settlement[.]”);
quoted in In re Worldcom, Inc. ERISA Litig., 339 F.
Supp. 2d 561, 567 (S.D.N.Y. 2004). In addition the litigants’
interests, the interests of potential competitors, the public at large,
and other third parties will be affected by this proposed settlement. If
the proposed settlement goes forward, Google will in many cases receive
preferential, if not sole, access to licenses for the books of both known
and orphan authors, with potentially harmful effects on consumers and
competitors. These effects may lead to violations of the Sherman
Antitrust Act, requiring the Court to reject the settlement as proposed.
See, e.g., Grunin v. Int’l House of Pancakes, 513 F.2d
114, 123 (8th Cir. 1975) (“a court cannot lend its approval to any
contract or agreement that violates the antitrust laws.”); In
re Montgomery County Real Estate Antitrust Litig., 83 F.R.D. 305,
319 (D. Md. 1979) (“no court may lend its approval to an illegal
agreement or to one which has an illegal effect.”). However, the
Court’s analysis of competitive effects does not start or end with
the letter of that law. The Court’s obligations are not merely to
prevent explicitly illegal acts in the settlement of a lawsuit, but also
to protect the public interest from other harms.

A. Google will become the sole party able to license orphan books.

Orphan works simply cannot be licensed due to the absence of their
copyright owners. In this settlement, however, Google proposes to
incorporate all registered orphan books into its licensing scheme absent
any ability of the absent authors to say otherwise.

As a sub-class of the proposed plaintiff class, orphan rightsholders
occupy a distinct legal position. Although some works may become
“un-orphaned” as rightsholders emerge, works are not
considered orphaned just because their rightsholders fail to find out
about their usage, nor simply because rightsholders lack a meager
financial incentive to identify themselves. Instead, many rightsholders
are unfindable because they do not know that they are the proper
holders of the rights. There are a number of reasons that a work might be
orphaned: a deceased author’s beneficiaries may not have been
properly notified that they had an ownership interest in a work, or even
be aware that a deceased author had created a work; the copyright owner
may not have updated the registration; transfer of ownership may not have
been properly recorded; or the author may have died without heirs,
reverting the copyright to the state. See U.S. Copyright Office,
Report on Orphan Works 21-29, available at http://www.copyright.gov/orphan/orphan-report-full.pdf.

Nor will the BRR or active rightsholders have much of an incentive to
alter this state of affairs. Since unclaimed funds for orphan works will,
after a time, be distributed among registered rightsholders or used to
pay BRR expenses, neither the BRR nor the registered rightsholders will
have a financial incentive to encourage orphans to emerge. See
Proposed Settlement §6.3(a).

This selective licensing of the orphan authors’ rights goes far
beyond the standard power of a class action suit to waive the claims of
non-present parties. As discussed below, orphan rightsholders are not
adequately represented by class counsel or named plaintiffs, and the
claims that absent rightsholders might have against Google for scanning
works stem from a different factual predicate than any claims they might
have for Google making their works available online. While courts
reviewing class action suits and settlements may make certain
determinations regarding parties not before the court, the procedural
requirements and safeguards surrounding the process prevent the court
from usurping the powers of the legislature. Yet the proposed settlement
shifts significant issues of copyright policy from the legislative
process into this particular lawsuit, deciding that a smaller set of
exclusive rights will be granted to the authors of certain books unless
they take affirmative action. Those who do not so act (or who cannot act)
are compelled to grant a license for the use of their works.

One of the consequences of this is that the landscape of copyright
changes, but only for Google and for the BRR. This results in a market
for orphan works that becomes the sole province of these two entities.
Anyone wishing to enter that market would have to risk extensive
copyright liability of a sort no one—including the
defendant—ever has. The competitive issues surrounding the
parties’ ability to use orphan works must therefore be examined
closely.

1. Google’s acquisition of a monopoly on orphan works violates
Sections 1 and 2 of the Sherman Antitrust Act

The settlement agreement constitutes a contract that grants Google the
sole right to license orphan (and otherwise unclaimed) works. This
agreement thus operates alongside copyright law to restrain trade and
maintain Google’s monopoly position unfairly, in violation of
Sections 1 and 2 of the Sherman Act, respectively.

Section 1 of the Sherman Act prohibits “every contract,
combination…or conspiracy in restraint of trade or commerce among
the several States, or with foreign nations.” 15U.S.C. §1.
This requires both the existence of a contract, combination, or
conspiracy, and a resulting restraint of trade. Standard Oil Co. v.
United States, 221 U.S. 1, 59-60 (1911).

Section 2 prohibits monopolization and any attempts to monopolize.
15U.S.C. §2. Violations of Section 2 require a finding of “(1)
the possession of monopoly power in the relevant market and (2) the
willful acquisition or maintenance of that power as distinguished from
growth or development as a consequence of a superior product, business
acumen, or historical accident.” Eastman Kodak Co. v. Image
Technical Servs., Inc., 504 U.S. 451, 480 (1992) (quoting United
States v. Grinnell Corp., 384 U.S. 563, 570-71 (1966)). While
Section 2 thus allows for monopolies to exist, exclusionary behavior on
the part of the monopolist is prohibited. Exclusionary behavior has been
defined as “behavior that not only (1) tends to impair the
opportunities of rivals, but also (2) either does not further competition
on the merits or does so in an unnecessarily restrictive way.”
Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585,
605 (1985) (quoting 3 P. Areeda & D. Turner, Antitrust Law
78 (1978)).

Since no party other than Google can license the use of orphan works,
Google will have an absolute monopoly on selling access to these works.
The agreement prevents Google from licensing to others the use of any of
the scanned works (Proposed Settlement §2.2), and unless
the agreement allows the BRR to license orphan works to other parties,
this means no other entity has the legal ability to display or distribute
orphan works.While the number of orphan books at stake may be
debated,^[2] it
remains true that for every single work orphaned, Google becomes the only
permitted user, insulated from potentially massive copyright liability.

Therefore, the settlement agreement that places Google in its monopoly
position violates Section 1 of the Sherman Act by serving as a contract,
combination, or conspiracy in restraint of trade. Through its agreement
with the plaintiff class, Google is able to circumvent a legal restraint
that applies to all other potential licensees. These other licensees, by
abiding by the copyright laws, will be laboring under the need to locate
potentially unlocatable orphan rightsholders before licensing their
works, and will therefore be restrained from even entering the
market.^[3]

Through this same sidestepping of legal barriers, the settlement
agreement allows Google to monopolize illegally in violation of Section 2
of the Sherman Act. Through the settlement, Google’s monopoly over
orphan works becomes an insurmountable one, maintaining monopoly by means
not attributable to competition in the marketplace. If Google is allowed
this ability to license orphan works, the only way for a competitor to
enter the market would be to (1) get sued by a class of authors broad
enough to include orphan authors, and (2) come to a settlement agreement
with that class to allow the provision of orphan works to the public.
Each of these situations depends upon a set of circumstances so unlikely
to occur as to be nearly impossible.

First of all, there are already substantial doubts as to whether a class
of plaintiffs suing a hypothetical Google competitor can actually be
approved if it is drawn so inclusively as to include orphan authors. As
noted above, such a broad class will likely fail to be adequately
represented by class counsel. As for a settlement agreement that
parallels the present one, plaintiffs would be unlikely to accede to an
agreement that would give them no additional advantages. A new licensing
scheme with this hypothetical defendant would not be allowed to contain
better terms than the agreement with Google, given the proposed
settlement’s “Most Favored Nation” (“MFN”)
clause. With Google already providing a market for digital provision of
books and little advantage to be gained by having an additional licensee,
the plaintiff class, if it ever formed, would likely seek a remedy that
would look substantially different from this settlement, and one that may
well not include a licensing provision. If indeed such a provision were
the aim of the hypothetical defendant, plaintiffs would also have every
incentive to advantage themselves in other provisions of the settlement,
such as monetary compensation. This likelihood would raise the cost and
risk of an already risky strategy for the defendant. All of these factors
together make it fanciful to suggest that competitors might enter the
market through engineering class action litigation.

Even if the plaintiff class and the hypothetical defendant were to come
to an agreement that mirrored the present proposed settlement, such an
outcome, unreachable via the remedies available in copyright law and
already demonstrated by this litigation, would suggest improper collusion
between the parties. As here, the future defendant would receive the
right to scan and display orphan works, a right unavailable to it through
the normal process of private contracting, while the named plaintiffs
would be reaping the benefits of waiving someone else’s rights for
them. But unlike here, both parties would already be aware of how class
action law had been used to strike this otherwise impossible bargain.
This Court should be wary of creating a market that can only be entered
through mass litigation.

Though Google will be able to offer a new product in a market heretofore
unavailable, this situation differs significantly from the monopoly
created by the inventor of a new business method or the holder of an
enforceable patent or copyright. Here, the barrier to competitors was
neither legislatively ordained (as with intellectual property
rightsholders), nor the result of historical happenstance (as with
monopolists who are business innovators). Instead, competitors are kept
from this market through the operation of copyright law and an
overextension of the factual predicate for the proposed releases. Unlike
new markets created through innovation and invention, the market in
orphan works will suffer from the lack of experimentation and negotiation
that accompany the development of new markets. Unlike a monopoly created
through the grant of a specific patent or copyright, there will be no
bidding for licenses for further use, nor the ability to invent around
the monopoly grant. Not only does the settlement attempt to use this
Court to create a new market, it asks the Court to now designate that
market’s structure and operation, naming the parties as the
dominant players. Absent an approval of a deal like the one proposed by
the settlement, the market would be an illegal one.

By permitting Google to enter a market legally barred to all other
competitors, the proposed settlement would restrain trade and allow
Google to monopolize the market in orphan books. This Court should not
only prohibit the parties from engaging in anticompetitive behavior, it
should refrain from acting as the means by which a monopoly is created
and maintained.

2. The Book Rights Registry should be subject to antitrust scrutiny.

Google is not the only entity that will amass market power via the
proposed settlement; the BRR itself also poses a significant antitrust
concern. Just as Google becomes the sole licensee of orphan books, the
BRR, created as a means by which to carry out the will of the plaintiffs
into the future, will become the assumed representative of the missing
authors. While active individual rightsholders can dilute the market
power of the BRR by arranging side licenses with distributors and
digitizers, orphan works—whose rightsholders are not available to
make such alternative licensing agreements—will be licensable only
by the BRR. This not only raises many of the same antitrust questions as
Google’s orphan works monopoly, it also leaves the BRR near
complete freedom to engage in price-fixing. Indeed, as prices are by
default set by the settlement agreement, subject to alteration by
individual copyright holders, where those copyright holders are unable to
make any decision, the prices will default to those specified in the
system agreed upon between Google and the BRR. See Proposed
Settlement §4.2(b)(i).

The status of orphan works is central to the proper outcome of this
litigation. Absent any alteration to the settlement agreement, the result
is the creation and maintenance of a new monopoly in orphan works,
enforced in part by this Court and in part by copyright law.

B. The settlement also violates the Sherman Act with regard to non-orphan
works.

The antitrust implications of the settlement agreement are not restricted
to orphan works; the agreement, as it stands, also has powerful market
effects on the availability of non-orphan works, whether they are
unclaimed or claimed.

1. The opt-out nature of the license unfairly advantages Google in
licensing the works of known authors

The anticompetitive facets of the settlement agreement are not limited to
orphan works. As noted above, the proposed settlement is structured so as
to convert the class action opt-out procedure into a quasi-statutory
licensing scheme. This legal alchemy would place the BRR in a singular
position of retaining, by default, the digital licensing rights to all
in-copyright registered U.S. books published before January
2009—even works for which direct licensing is not a practical or
possible option. See Proposed Settlement §1.16.
Beyond the difficulties that this creates for orphan works, the
settlement’s peculiarities also give Google an outsized market
advantage in licenses for works with known rightsholders.

It should first be noted that the market for these works will be
marginally more competitive than that for orphan works. Authors of known
works may choose their level of participation in the settlement and in
the ensuing licensing program. Furthermore, the license granted by the
authors to Google will be non-exclusive, allowing authors to license
their works to other electronic distributors, or even to alternative
licensing pools competing with the BRR.

However, the license to Google possesses an overwhelming advantage over
any other agreement that an author might attempt to enter into. Since all
class members are assumed to be included in the settlement, only those
who actively decide to opt out will be absent from the pool of licenses
available to Google. Any other licensee, who did not have the benefit of
the class action structure, would need the active approval of each author
it wished to include in its pool of licenses. Put another way, because
individual licensing is far harder than licensing through the BRR, only
the defendant will have effective access to all unclaimed works,
even those that are not truly orphaned.

The same would be true if the BRR itself, acting on behalf of
rightsholders, were to attempt to license works to any other entity.
While the proposed settlement contemplates the BRR potentially acting as
a rights clearinghouse for other parties (see Proposed
Settlement §6.2(b)(iii)), it is doubtful that the BRR would be
authorized to represent all class members in negotiations with entities
not party to this litigation.

This discrepancy in the pool of licenses available to Google, versus
others, creates a considerable, artificial market advantage. Even among
the smaller set of rightsholders who exist, know their rights, and have
been informed of the settlement, it is highly likely that far more
rightsholders will passively choose not to opt out of this arrangement
with Google than would actively opt into an agreement with any other
party. These rightsholders would then have, through inaction, granted an
ongoing, essentially exclusive license to Google. The proposed settlement
would thus give Google a sizeable lead in the novel market for the
collective licensing and display of digitized books.

This market is so novel as to be created by the parties’ very
actions in this litigation.^[4] Google’s proposed offering differs in
significant part from the existing market in digital books. Whereas
current digital book outlets such as Amazon’s Kindle store, Barnes
& Noble’s eBook store, and Google’s own current Book
Search product must build their offerings by obtaining rights from each
author to be included in their collections, Google’s use of the
opt-out class action mechanism means the majority of existing U.S. books
will automatically be on offer. The nature of this arrangement makes
Google not just quantitatively different from any potential competitors
in the number of books available, but qualitatively different. Access to
an entire body of digital books is valuable beyond the value of the
individual books combined, provided that the catalog is large enough. The
settlement thus creates a new sort of service—not merely access to
discrete texts, but access to a whole body of copyrighted literature. It
makes Google a “one-stop shop” for those who wish to find or
purchase books, because no other vendor can ever approach the level of
coverage that Google will possess.

This makes Google the most powerful, if not the sole, licensee of this
corpus and the sole purveyor of the corpus to individuals and
institutions alike. The settlement agreement could thus run afoul of
Section 2 of the Sherman Act, as Google will be exercising monopoly power
in this new market and maintaining that power through means other than
standard competition. Those means include the judicially-approved
settlement agreement itself. The irony of this is that the mechanism of
the class action is invoked precisely because of the difficulty of
assembling groups of rightsholders to grant blanket licenses. The same
difficulty used to justify the process that creates this new market will
serve to exclude competitors.

As with its effects on orphan works, the agreement leverages the class
action opt-out structure to act as a barrier to entry for other potential
licensees that does not result from the characteristics of the product,
acumen in pursuing market competition, or historical accident. Instead,
the restriction against competitors results from a mischaracterization of
the factual predicate of this litigation. Should the BRR be precluded
from licensing to others the works of class members to all interested
parties on the same terms as it licenses them to Google, Google will have
successfully used this proceeding to maintain its monopoly position in
violation of Section 2.

These same features of the settlement are problematic under Section 1 of
the Sherman Act as well. The settlement agreement as a whole can also be
characterized as a contract, combination, or conspiracy between Google
and the collective plaintiff class that operates in restraint of trade.
Though a mere refusal to license to others may be permissible under
Section 1, see Buffalo Broad. Co., Inc. v. Am. Soc’y. of
Composers, Authors and Publishers, 744 F.2d 917 (2d Cir. 1984), the
present situation involves more than an agreement that simply fails to
include competitors. Here, the restraint of trade results from a
combination of (1) the advantage granted by the opt-out structure and (2)
the copyright laws that simply and logically forbid potential competitors
from assuming that authors have agreed to license to them absent
affirmative consent. The same advantage manufactured by the use of the
class action to maintain monopoly status is thus also used to restrain
trade.

2. The Book Rights Registry should be scrutinized for antitrust
violations for claimed works as well.

As the future representative of the named plaintiff’s interests,
the BRR represents a horizontal combination of rightsholders who,
separately, might well compete with each other in pricing licenses for
the right to digitize and display their works. “[A]greements among
competitors to fix prices on their individual goods or services are among
those concerted activities that the Court has held to be within the
per se [illegal] category.” Broad. Music, Inc. v.
Columbia Broad. Sys., Inc., 441 U.S. 1, 8 (1979). Instead of an open
market between individual authors and potential digitizers, there is
instead the BRR, a combination of collective licensing body and
collecting society, much in the way that the American Society of
Composers, Authors, and Publishers (“ASCAP”) or Broadcast
Music, Inc. (“BMI”) negotiate licenses and collect revenues
for the creators of musical works. While those organizations currently
operate legally, this Court is no doubt aware that their operation has
for the past several decades been contingent upon constant scrutiny.
Given the history of antitrust scrutiny leveled at musical collecting
societies, it is only natural that the BRR be examined, at the least
under the rule of reason. See Id. at 24.^[5] This Court’s inquiry into the
BRR’s effects should also be more searching due to the fact that
unlike the performance rights societies, the BRR would be specifically
created under the auspices of this Court, and not merely examined for
legality under the antitrust statutes.

As a combination of book copyright holders, the BRR must be carefully
scrutinized to ensure that it does not act to restrain trade in licenses
to digitize books. As with orphan works, the BRR is the collective body
that will be negotiating the prices set for non-orphan works that are
bundled into institutional subscriptions. See Proposed
Settlement §4.1. Known authors will have little ability, except
through representation in the BRR, to determine what prices are set for
the subscription licenses. As noted above, the market for Google
subscription service is so differentiated from the sale and licensing of
individual books that authors’ abilities to conduct side
negotiations will have little to no effect on the BRR’s
price-setting ability.

The problems with the settlement agreement that give rise to monopoly
power in orphan works thus also affect the market for works whose authors
are known. The anticompetitive effects on the market for these works
should also be carefully scrutinized.

3. The Most Favored Nation clause implicates sections 1 and 2 of the
Sherman Antitrust Act.

In addition to the inherent market advantages granted to Google and the
BRR by the structure of the settlement agreement, section 3.8(a) of the
Proposed Settlement imposes a “most favored nation”
clause.^[6] This
clause would prevent the BRR (or any other book licensing body) from
offering a Google competitor a lower licensing fee. While an innovative
product will naturally allow a business to be a legal monopoly, the MFN
clause acts as a contract to unfairly restrain trade, and attempt to
maintain monopoly status by foreclosing competition.

i. The MFN clause acts as a contract to restrain trade.

The MFN clause in the proposed agreement easily acts as a restraint of
trade in violation of Section 1 of the Sherman Act. MFN clauses have come
under increasing antitrust enforcement scrutiny, due to their potential
for suppressing competition. See Jonathan M. Jacobson,
Antitrust Law and Developments 148 (6th ed. 2007). Antitrust
enforcement authorities have, in several cases, used consent decrees to
bar the use or enforcement of MFN clauses. See, e.g., United
States v. Delta Dental Plan, 1995-1 Trade Cas. (CCH) ¶71,048
(D. Ariz. 1995); United States v. Medical Mutual of Ohio, 1999-1
Trade Cas. (CCH) ¶72,465 (N.D. Ohio 1999); United States v.
Delta Dental, 1997-2 Trade Cas. (CCH) ¶71, 860 (D.R.I. 1997);
United States v. Oregon Dental Serv., 1995-2 Trade Cas. (CCH)
¶71, 062 (N.D. Cal. 1995); United States v. Lykes Bros. S.S.
Co., 60 Fed. Reg. 52, 208 (DOJ Oct. 5 1995); RxCare of
Tenn, 121 FTC 762 (1996).Federal courts have also upheld the notion
that MFN clauses may be anticompetitive. In United States v. Delta
Dental of Rhode Island, a district court agreed with the Department
of Justice that it is possible for an MFN clause to trigger a violation
of Section 1. 943 F. Supp. 172, 174-5 (D.R.I. 1996). In that case, the
government characterized an MFN clause in a contract between an insurance
company and its care providers as a “concerted action” in
violation of the Sherman Act. Id. The clause also threatened a
restraint of trade, as it aimed not to lower prices for an insurance
firm’s customers, but to exclude potential rivals. Id. at
177. The court held that that these characterizations could survive a
motion to dismiss. Id.

In the settlement agreement, the concerted action between the BRR and
Google serves to restrain trade by requiring any book licensing
collective—even one arising as an alternative or a competitor to
the BRR—to offer any later-arising licensees the same or a worse
deal than Google’s. Google may thus, for the ten years specified in
the MFN, preserve its first mover advantage while removing a major
incentive for the BRR to deal with alternative digital distributors.

It should also be noted that the MFN purports to apply not only to the
BRR, but to “any substantially similar entity” organized by
rightsholders. Not only does the settlement disadvantage potential
licensees, it also forecloses an offer of competitive rates by any
collective licensing bodies that might arise as alternatives to the BRR.
Thus, any authors who fail to opt out of the settlement
entirely—even those authors who do not register with the BRR and do
not enter into the profit-sharing agreement—will be bound by this
clause even if they create a separate collective to compete with the BRR.
In effect, the MFN makes such attempts at competition with the BRR
futile, unless the competing registry is to consist entirely of the
limited number of rightsholders who opt out of the whole settlement.

ii. The MFN clause serves as an attempt to foreclose competition.

The MFN clause also acts as an attempt at monopolization in violation of
Section 2 of the Sherman Act. Google’s position as a monopoly power
in the relevant market for mass digital access to books is clear.

The MFN clause maintains that power beyond what Google would be normally
entitled to based on its innovation in creating its digital book
products. The MFN clause artificially heightens existing barriers to
entry into the market for digital access to books. It also reduces
incentives for any other competitors to attempt to deal with the BRR or
any similarly situated licensing collective, ensuring that consumers will
have fewer choices for access to these works. MFN clauses are typically
allowed when they can be seen as a pro-competitive means for lowering the
costs passed on to consumers. Here, however, the absence of any existing
competition makes the MFN clause function far more like an exclusionary
measure than as a cost-reducing measure.

iii. The prerequisites for triggering the MFN clause do not save it from
scrutiny.

The MFN clause in the settlement agreement contains two conditions that
must be met before its restrictions are triggered. However, their
presence does not prevent the MFN clause from operating to restrain trade
or exclude competition.

The first condition—that less than ten years have elapsed from the
time of the Effective Date—merely allows the MFN provision to
sunset after a lengthy period of time. Ten years is an eternity in the
context of online, digital access to information—over a third the
lifetime of the commercially-available Internet.

The second condition requires the license to include a
“significant” number of rightsholders who have not registered
with the Registry. Although this limitation means that the MFN clause
will only apply to certain competitors, its nature ensures that only
those competitors who pose a more serious threat to Google’s
dominance—by offering books that are unavailable to
Google—will be targeted by the MFN. Whether such an anticipated
competing agreement contains a “substantial” number of
non-Registered Rightsholders because it involves orphan works or
Rightsholders who have decided for other reasons not to actively register
with the BRR, the MFN clause’s structure still operates to ensure
that Google will retain an advantage not only among competing licensees,
but that that advantage will extend to other licensors competing with the
BRR.

II. A class action settlement is the improper mechanism by which to
determine the rights of orphan works authors with regard to a single
user.

The very fact that this settlement, if approved, will create an entirely
new market, resolve the rights of parties not present, allocate payment
for those rights to the present parties, and license uses which where
never contemplated prior to the suit should put the Court on alert that
this is not a traditional class action settlement. The proposed
settlement stands poised to effectively alter the rights of
authors—especially those rightsholders of orphan
works—indefinitely. As described above, the result will be a legal
regime in which only one party can lawfully offer a large library of
books to the public. It accomplishes this through two legally
impermissible class action mechanisms. First, even though to date Google
has engaged only in the non-infringing scanning, indexing, and excerpting
of books, the proposed settlement purports to release it from liability
for the entirely new activities of full-text display and sale of those
books. Second, the provisional class includes many orphan rightsholders
who are not only poorly represented by the named plaintiffs, but have
interests antagonistic to those professing to represent them.

The antitrust problems described above illustrate exactly why class
action law does not permit releases of this type. If the law were to
allow a class, represented by only a few named plaintiffs, to go beyond
redress of harms and license entirely new activities, then parties will
have the incentive to use class actions as an easy mass licensing tool
rather than a method for aggregating and redressing group harm. With this
tool, named plaintiffs will be able to benefit from the mass licensing of
others’ rights without any actual pre-existing harm, proper notice,
adequate representation, or consent of absent class members. Such suits
will almost inevitably create monopolies because, as here, other parties
will not be able to directly license rights from the entire class, and
replicating a class action lawsuit as a defendant would be difficult or
impossible. It is difficult to see where the limit of such a class action
mechanism would lie.

The proper place both for changes to the defaults of copyright laws and
the balancing of interests of parties who are unable to protect their own
interests in court is the legislature, not class action law. The Court
should therefore not approve a settlement that shifts the balance of
copyright in favor of a lone party by mass licensing new uses of rights
belonging to an inadequately represented subclass of artists to create a
single point of access and distribution for their works.

A. The proposed settlement exceeds the scope of a permissible class
action release because it releases new, future activities which were not
part of the underlying suit.

Because the defendant has never sold or otherwise granted full-text
access to books without first obtaining a license, a settlement that
releases them for liability for these activities is impermissible under
Second Circuit law.

1. Google has never engaged in the unlicensed sale of the full text of
books, while the proposed settlement would license that use.

The proposed settlement licenses works—including unclaimed
works—for institutional and individual sale, allowing Google to
provide access to the entire text of those books. See, e.g.,
Proposed Settlement §3.3(a). The complaint, however,
contemplates only the “public display of portions of such
Books and Inserts on its commercial website.” Second Amended
Complaint at 14,Authors Guild et al. v. Google (S.D.N.Y. Oct.
30, 2008) (No. 05-CV-8136).^[7] In fact, Google has never provided and does not
currently provide the full text of works unless it has acquired a full
display license, and nothing in the record indicates otherwise.

The final notice of proposed settlement provides further evidence of this
disconnect between the scope of the claims and the scope of the
settlement. The notice describes only “claims that Google violated
the copyrights of authors, publishers and other owners of U.S. copyrights
in books and other writings by digitizing (scanning) them, creating an
electronic database of books, and displaying short excerpts
without the copyright owners’ permission.” Final Notice
of Class Action Settlement, available at
http://www.googlebooksettlement.com/intl/en/Final-Notice-of-Class-Action-Settlement.pdf
(emphasis added). And while it lists a “benefit” of the
settlement as a portion of the revenues from “sale of online access
to Books uses” and says that rightsholders may “determine
whether and to what extent Google may use their work,” there is
nothing in the notice to suggest that those uses will go far beyond the
“short excerpts” underlying the suit. See
id.^[8]

The settlement thus proposes to release future claims against the
defendant based on actions the defendant has never taken.

2. Class action law does not permit the release of behavior which was not
a part of the underlying suit and which has never occurred.

Because the Second Circuit does not allow releases of claims for future
activities which are not part of an ‘identical factual
predicate,’ much less those that have never occurred at all, the
Court cannot approve a settlement which authorizes these new activities
on behalf of class members.

In the Second Circuit, for a class action release to include claims not
presented, the conduct must “arise[] out of the ‘identical
factual predicate’ as the settled conduct.” Wal-Mart
Stores, Inc. v. Visa U.S.A., Inc., 396 F.3d 96, 107 (2d Cir. 2005).
That court has “previously ‘assumed that a settlement could
properly be framed so as to prevent class members from subsequently
asserting claims relying on a legal theory different from that relied
upon in the class action complaint but depending upon the very same
set of facts.’” TBK Partners, Ltd. v. W. Union
Corp., 675 F.2d 456, 460 (2d Cir. 1982) (emphasis added) (citing
National Super Spuds, Inc. v. New York Mercantile Exchange, 660
F.2d 9, 18 n.7 (2d Cir. 1981)). This condition is not met in the instant
case; plaintiffs could not have brought a claim for the offering of full
books based on the facts as they exist today. Such a claim was not and
could not have been presented because it is not part of an
“identical factual predicate” or “the very same set of
facts,” as indexing and display of excerpts, and therefore may not
be released.

In TBK Partners, the court explained the reason a court might
allow release of unpresented claims: “[I]n order to achieve a
comprehensive settlement that would prevent relitigation of settled
questions at the core of a class action, a court may permit the
release of a claim based on the identical factual predicate as that
underlying the claims in the settled class action even though the claim
was not presented and might not have been presentable in the class
action.” TBK Partners 460 (emphasis added). Were the
instant case to be litigated to completion, a later case involving claims
for the full text display of books would raise no settled questions.
There is no question raised by the current case regarding the full-text
display of books. The record points only to legal claims about scanning
and short excerpts of books. Even in listing questions common to the
class, the amended complaint raises no questions about whether Google had
engaged in the display or sale of the full text of books or whether such
activities constitute copyright infringement. See Second Amended
Complaint at 13-14, Authors Guild et al. v. Google (S.D.N.Y.
Oct. 30, 2008) (No. 05-CV-8136).

Were this suit to be pursued to completion, the only questions settled
would be whether defendant’s previous activities of scanning,
indexing, and displaying “short excerpts” were infringing. A
later suit alleging infringement for full display as well as consumer and
institutional sale of online book access would raise wholly different
legal questions and a determination of legality would have to be based on
specific factual circumstances—circumstances which could not have
been litigated in this action because they have not occurred. See,
e.g., Harper & Row Publishers, Inc. v. Nation
Enterprises, 471 U.S. 539 (1985) (explaining that in copyright
litigation, “fair use analysis must always be tailored to the
individual case”); H.R. Rep. No. 94-1476, at 65-66 (1976),
reprinted in 1976U.S.C.C.A.N. 5679 (“the endless variety
of situations and combinations of circumstances that can rise in
particular cases precludes the formulation of exact rules in the
statute…. the courts must be free to adapt the doctrine to
particular situations on a case-by-case basis”).

When the Second Circuit has approved settlements releasing additional
claims, they have been based on events which had already occurred, and
would have merely have provided alternate venues or formulations of the
claims. For instance, in Wal-Mart, the additional claims
released were alternate legal theories based on the same set of
exclusionary rules which credit card companies had been sued for using.
Wal-Mart 107-08. Similarly, in TBK Partners,
the court allowed the release of state court claims in a federal suit
because those claims “hinge[d] on the identical operative factual
predicate: the correct valuation of whatever reversionary interest was
owed to … shareholders.” TBK Partners 460.

Neither example approaches the factual disparity present here, as both
were based on different ways of litigating the same, pre-existing facts.
Offering the full text of a book is a different activity and would
provide a wholly different set of facts than scanning, indexing, or
presenting short “snippets” of books. Allowing purchase of
full access to those books, either individually by consumers or in bulk
by institutions is even farther from the events which have actually
occurred. None of these proposed uses are part of the “identical
factual predicate” as the existing uses.

The instant case also fails tests announced in other circuits but cited
for support by the Second Circuit. See, e.g., Wal-Mart
460 n.13 (“The Fifth Circuit has noted, ‘The weight of
authority establishes that … a court may release not only those
claims alleged in the complaint and before the court, but also claims
which could have been alleged by reason of or in connection with any
matter or fact set forth or referred to in the
complaint.’”) (quoting In re Corrugated Container
Antitrust Litig., 643 F.2d 195, 221 (5th Cir. Apr.1981) (emphasis
added)). And as discussed above, none of the potential facts including
sale and display of full works were “set forth or referred to in
the complaint.” This Court therefore may not release those
additional claims here, licensing future uses of orphan works which were
never described in the complaint or executed in fact.

3. If the Court concludes that the released claims share the same factual
predicate, it should require the release of other potential defendants.

It is worth noting that if the Court concludes that the release and the
complaint do share the same factual predicate, then the Court may also be
able to approve a settlement which releases non-parties from liability.
See Wal-Mart 109 (approving the release of “claims against
non-parties where … the claims against the non-party being released
were based on the same underlying factual predicate as the claims
asserted against the parties to the action being settled” when
released non-parties were corporate members of defendants’
organization and contributed to the settlement.).

Allowing other providers to license orphan and unclaimed works under the
same or similar terms as the defendant will result in broader access to
books and the reduction of the concerns surrounding a single party being
the only legally protected source for those books. Such an outcome, while
still problematic, would be far better for the public good and the goals
of competitive, open access to creative works and protection of orphan
authors from the will of a single distributor, and would prove less
troubling than the settlement as it stands today. Therefore, if the Court
finds that offering the full text of books falls within the same factual
predicate as scanning, indexing, and excerpts, it should only approve a
settlement which releases other potential users of plaintiffs’
books from liability under the same or similar or similar terms.

B. The provisional plaintiff class can not be certified because orphan
authors’ interests conflict with named plaintiffs’ in
violation of F.R.C.P. 23(a)(4).

The rights of orphan authors are not well-protected when the named
plaintiffs, plaintiff organizations, and defendant all stand to directly
and continually benefit from the licensing of rights belonging to others.

Because the provisional class contains a significant subset of
“orphan rightsholder” members who have interests antagonistic
to those of the named plaintiffs, the class does not meet the statutory
requirements of F.R.C.P. 23(a)(4). Rule 23(a)(4) requires that “the
representative parties will fairly and adequately protect the interests
of the class” before a class may be certified. The Second Circuit
has “provisionally certified for settlement purposes only” a
class comprising “All Persons that, as of January 5, 2009, have a
Copyright Interest in one or more Books or Inserts.” Order Granting
Preliminary Settlement Approval at 2, Authors Guild et al. v.
Google (S.D.N.Y. Nov. 11, 2008) (No. 05-CV-8136).

As described above, this class contains an unknown, but significant
number of “orphan rightsholders”—owners of copyright
interests who are unidentified and potentially unidentifiable. Named
plaintiffs, on the other hand, comprise several individual, non-orphan
authors, as well as membership organizations covering a very small
fraction of book rightsholders. These named plaintiffs stand to
continually financially benefit from licensing of the rights belonging to
owners of orphans and unclaimed works. Because there is a large subclass
with fundamentally different interests from the named plaintiffs and
because the presence of these class members undermines the typicality and
commonality requirements for a class certification, a class containing
orphan rightsholders should not be certified.

To ensure adequate representation under F.R.C.P. 23(a)(4), “a
district court must determine whether plaintiff’s interests are
antagonistic to the interest of other members of the class.”
Cent. States Southeast and Sw. Areas Health and Welfare Fund v.
Merck-Medco Managed Care, L.L.C., 504 F.3d 229, 245 (2d Cir. 2007)
(internal quotation omitted). Here, orphan rightsholders have interests
directly opposed to those of named plaintiffs, who have updated their
registrations, are represented by the Authors Guild or the Association of
American Publishers, are actively licensing and exploiting their works,
or have claimed their works directly, especially in light of the revenue
structure dictated by the proposed settlement.

In the proposed settlement, rightsholders who register with the BRR
(“registered rightsholders”) will receive both a portion of
the revenues from subscriptions (which include access to unclaimed works)
and a portion of the revenues directly attributable to both sale of and
the advertising placed around them. See Proposed Settlement
§6.3(a). Thus, registered rightsholders and named plaintiffs will
financially benefit if the orphan works remain orphaned, and if the class
members who are orphan rightsholders are not found and do not claim their
works, either before or after the settlement is approved. And unlike most
class actions, where compensation is either shared at the time of
settlement or dispersed from a fund for later-discovered harms, this fund
contains a continuing incentive for current beneficiaries to reduce the
compensation to later ones. Because those who register with the BRR will
continue to receive payments for use of unclaimed works, they will
benefit indefinitely from those books remaining unclaimed. Orphan
authors, however, are in the opposite situation, and will receive no
benefit at all unless they are located and given the chance to exercise
their rights and claim compensation.

A unitary class such as this one, with conflicting subclasses, is
therefore improper. For example, in Amchem Products v. Windsor,
the Supreme Court analyzed a proposed settlement for asbestos exposure to
determine, inter alia, whether class members who are already
injured had sufficiently overlapping interests with those who were merely
exposed to meet the requirements of Rule 24(a)(4). Amchem Products,
Inc. v. Windsor, 521 U.S. 591, 626-28 (1997). The Court concluded
that because there were “discrete subclasses” with
conflicting interests, a settlement could not be approved without
separate subclasses with corresponding representatives. Id. 627.
Because “for the currently injured, the critical goal [was]
generous immediate payments” while “the interest of
exposure-only plaintiffs [was] ensuring an ample, inflation-protected
fund for the future,” a unitary class was improper. Id.
627.

The conflict of interest present in the instant case is even more direct.
In Amchem, as in most class actions, both the immediate payment
and the fund could be evaluated at the time of settlement. Those already
injured would receive payment immediately, and would not have future
incentives to reduce the fund’s size. Those governing the fund
would not receive any benefit if those who later discovered harms stepped
forward to claim their share. Here, the opposite is true. Although the
proposed settlement proposes that the BRR “will attempt to locate
Rightsholders with respect to Books and Inserts,” Proposed
Settlement §6.1(c), the author and publisher rightsholders who
govern the BRR (see Proposed Settlement §6.2(b))
will have a continuing, perpetual, and opposite incentive not to
find orphan rightsholders who would then claim a portion of the fund.

Cases where the Second Circuit has found adequate representation when two
different subclasses are inapposite. For instance, in In re Visa
Check/MasterMoney Antitrust Litigation, 280 F.3d 124 (2d Cir. 2001)
(superseded by statute on other grounds), the court found that where
methods of calculating damages would result in differing recovery amounts
for plaintiff class members, interests were not sufficiently antagonistic
to deny certification. Here, however, the conflict of interest goes far
beyond an initial allocation of money, and is neither
“speculative” nor “hypothetical,” but fundamental
to the treatment of the orphan rightsholder subclass. Cf.
id. 145.

Orphan rightsholders’ conflicting interests go far beyond purely
monetary concerns. Named plaintiffs and other rightsholders who are
actively engaged in the licensing and sale of their works gain benefits
and abilities under the settlement that orphan rightsholders effectively
do not. Long after the chance to opt out of the settlement has passed,
named plaintiffs will be able, through the BRR, to control the price,
terms of use, and even availability of their works. Owners of orphan
works, by definition, cannot do any of these things. This means that
orphan works will continue to be made available perpetually under default
terms while those defaults are determined by those who can rest safe in
the knowledge that they may change them in the future. In essence, the
fate of orphan works and the rights of their owners will be permanently
at the mercy of the large publishers, Authors Guild members, and the lone
distributor who govern the BRR.

This conflict of interest is problematic for other class certification
requirements, as well. “The adequacy-of-representation requirement
‘tend[s] to merge’ with the commonality and typicality
criteria of Rule 23(a), which ‘serve as guideposts for determining
whether … maintenance of a class action is economical and whether
the named plaintiff’s claim and the class claims are so
interrelated that the interests of the class members will be fairly and
adequately protected in their absence.’” Amchem
Products 626 n.20 (quoting General Telephone Co. of Southwest v.
Falcon, 457 U.S. 147, 157 n. 13 (1982)).

For the reasons given above, rights of orphan authors are neither fairly
nor adequately represented by named plaintiffs who are able to later
control the use of their works, restrict the use of their works, and
financially benefit from the absence of those orphan rightsholders. As
the Second Circuit has observed, even if class representatives believe
“that the Settlement serves the aggregate interests of the entire
class, [] the adversity among subgroups requires that the members of each
subgroup cannot be bound to a settlement except by consent given by those
who understand that their role is to represent solely the members of
their respective subgroups.” In re Joint Eastern and S. Dist.
Asbestos Litig., 982 F.2d 721, 743 (2d Cir. 1992). Therefore, a
singular class with the named representatives, whose interests are
antagonistic to orphan rightsholders, should not be certified for this
proposed settlement.

C. The settlement is unnecessary and incompatible with the purpose of
class actions.

The proposed release of new, unrelated claims runs afoul of the basic
purpose of class action law: to aggregate individual claims for past
harms. Neither registered nor unregistered rightsholders need this class
action settlement to collectively redress past copyright harms, because
to date no such harms have occurred. Nor is a class action necessary to
enable the future, licensed full-text use of the works of those
rightsholders who have made themselves available. Even in the absence of
this settlement, Google could fund a Books Rights Registry and license
future uses of those rightsholders’ works directly. Structured
correctly to be open and competitive, such a registry would likely pass
antitrust scrutiny. The only part of this settlement which could
not be accomplished without the aid of a class action is the licensing of
future uses of works belonging to rightsholders who are not present to
defend their rights, as they have chosen not to register with the BRR or
are altogether inaccessible.

“‘The policy at the very core of the class action mechanism
is to overcome the problem that small recoveries do not provide the
incentive for any individual to bring a solo action prosecuting his or
her rights. A class action solves this problem by aggregating the
relatively paltry potential recoveries into something worth
someone’s (usually an attorney’s) labor.’”
Amchem Products 617 (quoting Mace v. Van Ru Credit
Corp., 109 F.3d 338, 344 (1997)). The problem here is not that
orphan or unregistered rightsholders lack the incentives to bring solo
actions. First, there is as yet no harm for them to seek redress as no
infringing uses of any works have been made. Second, with the exception
of a 5-year window within which they may be able to claim collected
royalties, these people will continue to receive no redress. But unlike
most class actions, where this failure to receive compensation for past
harms is balanced by the judicial finality and closure of the issue, this
settlement contemplates indefinite future uses (and legal
“harms”), making ongoing use of the rights that belong to
orphan rightsholders without any real likelihood of compensation for
those future uses.

Because the basic purpose of class action law does not support the
licensing of new, future uses of underrepresented parties’ rights,
the Court should not approve the perpetual licensing of orphan
authors’ rights through this proposed settlement.

III. The Court should perform additional review before approving a
settlement with the above problems.

A. The Court should deny class certification and not approve a settlement
which distributes the rights of orphan authors to a single party.

A number of the described problems with the proposed settlement and
provisional class certification stem from the inclusion of orphan
rightsholders in the settlement. Orphan rightsholders’ interests
are not represented by named plaintiffs, who are actively exploiting
their own rights. Further, the distribution of these rights without the
permission of the owners and only to a single defendant amounts to a
wholesale change to the way copyright law is applied to the defendant
with regard to books. This type of broad change to the copyright
landscape can and should only be made by the federal legislature, which
defines copyright law in the first place, and which has the institutional
expertise to both balance the interests of the public and the various
stakeholders and create a legitimate, nationwide solution which does not
create a new monopoly. Therefore, the simplest solution with regard to
the settlement is to deny certification to any class which includes the
orphan rightsholders. This may mean that no opt-out class of copyright
holders can be certified, or that such a class must be carefully
constructed to only include the works belonging to known, identifiable
rightsholders.

The Court can also remedy this problem by the equivalent action of
leaving the class description broad, but making the class opt-in. In this
manner, the settlement will only affect those who have chosen to allocate
their rights in this way, effectively removing unclaimed works and their
rightsholders from the settlement. The settlement would then remain as a
far more legitimate way to settle the future rights of those who actively
want their default rights under copyright law changed in this fashion.

B. The Court should seek to allow any competitor to license under the
same terms as Google.

There are two ways in which the Court might improve competition: remove
the MFN clause, and expand the settlement to allow licensing to third
parties.

1. Eliminate or alter the MFN clause.

The MFN clause presents one of the more easily-solved problems of the
settlement. Its removal from the settlement would eliminate any
anticompetitive effects. Alternatively, the MFN could simply be made
symmetrical, so that no other licensee would be faced with an additional
disadvantage to Google as a competitor.

2. Allow competitors to obtain licenses to digitize books.

Even removal or alteration of the MFN clause does not guarantee that
potential Google competitors will be able to competitively license the
display of books from the BRR. Although the settlement contemplates
licenses from the BRR to other parties, there is no indication that
either this settlement or copyright law would permit the BRR to issue
licenses for a list of books constructed via an opt-out procedure, as it
will for Google. Instead, other licensees would only be able to receive
licenses for the smaller set of books whose authors have opted in to a
licensing agreement. See Randal C. Picker, The Google Book
Search Settlement: A New Orphan-Works Monopoly?, John M. Olin L.
& Econ. Working Paper No. 462 at 14 (2009), available at
http://www.law.uchicago.edu/Lawecon/index.html;
Tom Krazit, Google Pushes for New Law on Orphan Books, Cnet
(July 31, 2009), at http://news.cnet.com/8301-1023_3-10300887-93.html.

It is this disparity that creates the antitrust and overall policy
questions that are most troubling with the proposed outcome of the
settlement. To eliminate this disparity, the settlement agreement should
ensure that Google must license works to other book access providers on
the same or similar terms as it itself receives. Such an arrangement
would simply require the plaintiff class to grant Google an additional
ability, and require Google not to abuse its market power. That power
should be disciplined by antitrust scrutiny and the threat of antitrust
litigation.

A more complete solution to the competition problem would be to allow the
BRR to license to other parties the same sets of works that are available
to Google. If the factual predicate on which the plaintiffs’ claims
are based encompasses behavior such as the display of full text by
Google, then it may be sufficiently broad as to encompass similar uses by
third parties. See supra at 28. In such a scenario,
Google would be able to provide access to its corpus of scanned works,
while other digitizers would also be able to compete with Google in the
new market for mass access to books. Existing collections of digitized
books could be made available through subscription services by acquiring
the appropriate license from the BRR.

C. The Court should consider input from relevant agencies.

The BRR, with or without the ability to license works on an opt-out
basis, would still be the most centralized clearinghouse of book rights,
and a powerful collective. Its role as a gatekeeper to bulk and blanket
licenses of books deserves the same ongoing supervision afforded to the
performing rights societies. The BRR should be monitored by the
appropriate antitrust authorities under a consent decree, to ensure that
licensee and licensors approaching it will be treated fairly, and not
solely to the benefit of incumbent parties in its licensing agreements.
The Court should also carefully consider the input of federal agencies
concerned with competition, consumer protection, and copyright law such
as the Department of Justice Antitrust Division, Federal Trade
Commission, and Copyright Office, and seek the input of such agencies
where proper.

CONCLUSION

The goal of access to written works is a noble one. However, access by
means of judicially-granted monopoly does not comport with the need for
access to information to be available through as many avenues as
possible. The current structure of the settlement agreement would permit
certain forms of access to orphan works through the single channel of
Google and the BRR, but truly open access must be granted not by grace,
but by right. For the foregoing reasons, the Court should therefore not
approve a settlement which creates this forward-looking single-source
license for unclaimed and orphan works.

In the alternative, a broad interpretation of the factual predicate
underlying the proposed settlement can justify not just the settlement as
proposed, but also a settlement that releases third parties from conduct
identical to Google’s. If this Court allows the release of claims
stemming from Google’s future offerings of full-text works, other
parties should be provided the same latitude to provide access to the
public, fueling a freer public discourse in a competitive landscape.

Jef Pearlman

Sherwin Siy

Public Knowledge

1875 Connecticut Ave., NW

Suite 650

Washington, DC 20009

(202) 518-0020

Sept. 8, 2009

^[1] We note that the
settlement itself does not specifically reference “orphan
works,” but divides works only into “claimed” and
“unclaimed” categories. Because it is effectively impossible
to individually distinguish between rightsholders who are unidentifiable
or nonexistent and those who have simply chosen not to participate, the
proposed settlement is poised to alter the default rules of copyright for
the owners of unclaimed worked as well as true orphans. The arguments
presented therefore apply with nearly equal force to those rightsholders.

^[2] A study by
Carnegie Mellon, based upon a random, statistically significant sample of
books in its collection, found that 22% of the time, the publishers of a
book could not be located. Of the books whose publishers could be
identified, 36% did not respond to letters of inquiry. This suggests that
up to 50% of books may have rightsholders who would not come forward.
Denise Troll Covey, Comments of Carnegie Mellon University Libraries
on Orphan Works (Mar. 22, 2005), available at http://www.copyright.gov/orphan/comments/OW0537-CarnegieMellon.pdf.

^[3] The
Noerr-Pennington doctrine, which immunizes from antitrust action
efforts to restrain trade through lobbying the government, does not
immunize all uses of the law to restrain trade or monopolize. If the
restraint results directly from private action, there is no immunity.
See Allied Tube & Conduit Corp. v. Indian Head, Inc., 486
U.S. 492, 499-500 (1988). Statutes themselves have been held in the
Second Circuit to be in furtherance of anticompetitive behavior. See
Freedom Holdings, Inc. v. Spitzer, 363 F.3d 149 (2d Cir. 2004). In
this case, the restraint of trade occurs through the private action of
the settlement agreement, which allows existing law to restrain others
from competing with the parties.

^[4] The fact that
this settlement agreement goes so far as to create entirely new types of
markets should serve as some indication that the process of class action
settlement is performing functions for which it was not designed. See
Part II, infra.

^[5] The licensing
scheme to be employed by the BRR can be compared to that used by the
performance rights societies in that just like ASCAP and BMI, the BRR
would offer licenses to a vast repertory of works. It is important to
recognize that this comparison overlooks one crucial difference: unlike
the BRR, ASCAP and BMI do not offer licenses for works whose copyright
owners are not their members, and membership is entirely opt-in. See
Michael B. Rutner, The ASCAP Licensing Model and the Internet: A
Potential Solution to High-Tech Copyright Infringement, 39 B.C. L.
Rev. 1061, 1074-1081 (1998)<>; Timothy Wu, Copyright’s
Communications Policy, 103 Mich. L. Rev. 278 (2004)).

^[6] Section 3.8(a)
states, in relevant part:

The Registry (and any substantially similar entity organized by
Rightsholders…) will extend economic and other terms to Google
that, when taken as a whole, do not disfavor or disadvantage Google as
compared to any other substantially similar authorizations granted to
third parties by the Registry (or any substantially similar entity
organized by Rightsholders…) when such authorizations (i) are made
within ten (10) years of the Effective Date and (ii) include rights
granted from a significant portion of Rightsholders other than
Registered Rightsholders…..

^[7] Notably, among
the amendments made to Plaintiffs’ complaint was the addition of
the word “portions” to its description of Google’s
display of works on its website. Cf. Amended Complaint
¶25(e), Authors Guild et al. v. Google (S.D.N.Y. Oct. 30,
2008) (No. 05-CV-8136); Second Amended Complaint ¶41(e), Authors
Guild et al. v. Google (S.D.N.Y. Oct. 30, 2008) (No. 05-CV-8136).

^[8] We have not
analyzed legally significance of the language in the notice, since when
the underlying settlement is impermissible, the notice is irrelevant.
However, the lack of clarity about the difference between the scope of
the settlement and the scope of the claims suggests that notice may be
inadequate.