The European Database Directive: Regional Stepping Stone to an International Model?

Fordham Intellectual Property, Media and Entertainment Law Fordham Intellectual Property, Media and Entertainment Law

Journal Journal

Volume 7

Volume VII

Number 2

Volume VII Book 2

Article 8

1997

The European Database Directive: Regional Stepping Stone to an The European Database Directive: Regional Stepping Stone to an

International Model? International Model?

G. M. Hunsucker

Phelps Dunbar L.L.P.

Follow this and additional works at: https://ir.lawnet.fordham.edu/iplj

Part of the Entertainment, Arts, and Sports Law Commons, and the Intellectual Property Law

Commons

Recommended Citation Recommended Citation

G. M. Hunsucker,

The European Database Directive: Regional Stepping Stone to an International Model?

, 7

Fordham Intell. Prop. Media & Ent. L.J. 697 (1997).

Available at: https://ir.lawnet.fordham.edu/iplj/vol7/iss2/8

This Article is brought to you for free and open access by FLASH: The Fordham Law Archive of Scholarship and

History. It has been accepted for inclusion in Fordham Intellectual Property, Media and Entertainment Law Journal

by an authorized editor of FLASH: The Fordham Law Archive of Scholarship and History. For more information,

please contact tmelnick@law.fordham.edu.

697

The European Database Directive:

Regional Stepping Stone to an

International Model?

G. M. Hunsucker

*

TABLE OF CONTENTS

INTRODUCTION............................................................................700

I. DATABASE PROTECTION UNDER EXISTING LEGAL

R

EGIMES.................................................................................710

A. General Characteristics of the Splintered Database

Markets ............................................................................711

B. Protection Afforded Database Makers by Existing

Legal Regimes ..................................................................714

1. Protection Afforded by Copyright.......................715

2. Protection Afforded by Contract..........................717

3. Protection Afforded by Unfair Competition

Law ...........................................................................720

II. C

OMPARISON OF THE DATABASE DIRECTIVE WITH THE

PROPOSALS ............................................................................722

* Copyright © 1997 by G. M. Hunsucker. Attorney-at-Law, Phelps Dunbar,

L.L.P., Tupelo, MS; Corporate/Executive Database Designer. Valdosta Univer-

sity, B.S.Ed 1992; Vanderbilt University School of Law, J.D. 1996. This Article

significantly expands the analysis set forth in a previous comment, Raising a Toll

Fence to Protect the “Noncreative” Labors of Database Makers: The European Database

Directive, awarded First Prize in 1996 Nathan Burkan Memorial National Compe-

tition at Vanderbilt University School of Law, sponsored by the American Soci-

ety of Composers, Authors, and Publishers. The author thanks Professor Jerome

H. Reichman for his comments on earlier drafts of the comment, and for discus-

sion of the issues raised by the sui generis regimes.

698 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

A. Scope of Protection...........................................................722

1. Database Defined ...................................................722

2. Sui Generis Rights Granted....................................724

a. The Extraction Right ........................................725

b. The Complementary Reutilization, or

Use and Reuse, Right.......................................725

c. Breadth of the Sui Generis Rights ...................726

3. Rights and Obligations in Existing Databases ...729

B. Beneficiaries of Protection................................................730

C. Term of Protection ...........................................................731

D. Exceptions: User Protections .........................................733

E. The Proposals’ Provisions Regarding Protection-

Defeating Devices and Database Management

Information ......................................................................734

III. MAKING THE PARADIGMATIC SHIFT: LINKING

INCENTIVES TO PRODUCTION COSTS ....................................735

A. Vindicating the Competitive Ethos in Nonproprietary

Data: Forcing Free Riders to Become Fair Followers.....740

1. The Facts of ProCD v. Zeidenberg..........................742

2. Protecting the Database Maker’s Investment

From Free Riders and Information Samaritans .743

3. Protecting the Public’s Interest in Competition.746

B. Criticisms of the Database Directive Regarding Sole-

Source Data: Is the Sky Falling?....................................749

C. The Ameliorative Measures: Will They Protect the

Public’s Interest in Competition? ...................................752

1. The Swords: The European Economic

Community Treaty and Competition Law.........752

2. Mandatory and Optional Shields to

Protect the User.......................................................755

a. The Mandatory Shields....................................756

b. The Optional “Fair Use” Shields ....................758

1997] EUROPEAN DATABASE DIRECTIVE 699

D. Effects within the European Community.........................761

IV. THE DATABASE DIRECTIVE: STEPPING STONE TO AN

INTERNATIONAL MODEL.......................................................763

A. Enabling Competition in a Commercialized

Information Market .........................................................763

1. Mandating Licenses in the Case of Sole-Source

Data ..........................................................................764

2. Compulsory Licenses Should Serve the

Interests of the Database Maker and the

Public........................................................................764

3. Prohibiting Exclusive Licensing

Arrangements .........................................................766

B. Potential Effects on Basic Research and Development....767

1. Concerns Expressed by Those Conducting

Basic Research.........................................................769

2. Potential Benefits to Those Conducting Basic

Research...................................................................771

a. Market Participation ........................................771

b. An International Treaty Addressing

Publicly-Funded Data......................................773

c. Reducing Inefficiencies Masked by

Hidden Subsidies .............................................774

d. Not a Threat to Technological Innovation....775

C. Potential Benefits to Society ............................................775

C

ONCLUSION ...............................................................................777

700 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

INTRODUCTION

Electronic databases

1

provide information users with ex-

traction tools enabling them to sort and arrange data in ways

meaningful to them; extending the manipulative abilities of

information users’ minds adds immense value to what

would otherwise be a mass of incoherent, disparate data.

2

1. The European Database Directive defines a database as “a collection of

works, data or other independent materials arranged in a systematic or methodi-

cal way and capable of being accessed by electronic or other means.” Council

Directive No. 96/9, art. 1(1), O.J. L 77/20, at 24 (1996) [hereinafter Database Di-

rective]. The Database Investment and Intellectual Property Antipiracy Act of

1996 (“U.S. Proposal”) defines a database as “a collection, assembly or compila-

tion, in any form or medium now or later known or developed, of works, data or

other materials, arranged in a systematic or methodical way.” H.R. REP. NO.

3531, 104th Cong., 2d Sess. § 2 (1996). The Draft Treaty on Intellectual Property

in Respect of Databases defines a database as “a collection of independent works,

data or other materials arranged in a systematic or methodical way and capable

of being accessed by electronic or other means.” Diplomatic Conference on Certain

Copyright and Neighboring Rights Questions, Draft Treaty on Intellectual Property in

Respect of Databases, art. 2(1) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996) [hereinaf-

ter WIPO Proposal]. Using the right collection tool, almost anything, including

numbers, text, audio, and images, can be reduced to binary form which can then

be compiled in a database. However, the U.S. and WIPO Proposal (“Proposals”)

exclude computer programs from their protection, and the Database Directive

excludes “computer programs used in the making or operation of databases ac-

cessible by electronic means” from its protection. See H.R. REP. NO. 3531, supra, §

3(d); WIPO Proposal, supra, art. 1(4) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996);

Database Directive, supra, art. 1(3), O.J. L 77/20, at 24 (1996). In addition, the Da-

tabase Directive and WIPO Proposal exclude certain dependent works. See Da-

tabase Directive, supra, recital 17, O.J. L 77/20, at 21 (1996); WIPO Proposal, supra,

Notes § 2.03 WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

2. The value of the global information industry could reach $3 trillion dol-

lars by the early 21st Century. Statement on Dismantling the U.S. Department of

Commerce: Testimony Before the Subcomm. on Government Management,

Information

and Technology of the House Comm. on Government Reform and Oversight, 104th

Cong., 1st Sess. (1995) (statement of Ronald H. Brown, Secretary of Commerce).

Note, however, that the “global information industry” referred to by the late Sec-

retary Brown is much broader than the database market alone. A recent state-

ment of the Co-Chair on High Performance Computing and Communications

valued the “information technology” sector at $500 billion. Prepared Statement on

Evolving the High Performance Computing and Communications Initiative to Support

the Nation’s Information Infrastructure: Testimony Before the Science Subcomm. on

Basic Research High Performance Computing and Communications Program of the

House Committee on High Performance Computing and Communications, 104th Cong.,

1st Sess. (1995) (prepared statement of Ivan E. Sutherland, Ph.D., Co-Chair,

1997] EUROPEAN DATABASE DIRECTIVE 701

Yet, in most of the world, these valuable business tools

3

stand virtually naked in the marketplace,

4

vulnerable to

Committee on High Performance Computing and Communications, and Vice-

President, Sun Microsystems). When the European Community (“EC”) penned

its Initial Proposal for the legal protection of databases in 1992, the EC database

market was estimated at $10.2 billion, which amounted to approximately 30% of

the world market. W. Joseph Melnik, A Comparative Analysis for the Legal Protec-

tion of Computerized Databases: NAFTA vs. The European Communities, 26 CASE W.

RES. J. INT’L L. 57, 59 n.14 (1994).

3. See supra note 2 (discussing the dollar value of the information technology

market); J. H. Reichman, Electronic Information Tools: The Outer Edge of World In-

tellectual Property Law, 17 DAYTON L. REV. 797, 806 n.45 (1992); Database Directive,

supra note 1, recital 9, O.J. L 77/20, at 20 (1996) (“databases are a vital tool in the

development of an information market within the Community”); WIPO Proposal,

supra note 1, Preamble WIPO Doc. CRNR/DC/6 (Aug. 30, 1996) (“databases

are . . . an essential tool for promoting economic, cultural, and technological ad-

vancement”); see also 142 CONG. REC. E890-04 (daily ed. May 23, 1996) (statement

of Sen. Moorhead) (“[D]atabases are an essential tool for improving productivity,

advancing education and training, . . . creating a more informed citizenry[,

and] . . . are also the linchpin of a dynamic commercial industry in the United

States.”).

4. Databases, like most forms of intellectual property, are saddled with

“public goods” characteristics: they are indivisible, inexhaustible, and ubiqui-

tous. See, e.g., J. H. Reichman, Charting the Collapse of the Patent-Copyright Dichot-

omy: Premises for a Restructured International Intellectual Property System, 13

CARDOZO ARTS & ENT. L.J. 475, 486 (1995); see also J. H. Reichman & Pamela Sam-

uelson, Intellectual Property Rights in Data?, 50 VAND. L. REV. 51, 59 n.35 (1997); see

infra notes 5, 8-10 and accompanying text (discussing vulnerability of databases).

Practically speaking, this means that without some legal protection, parasitic

competitors can access, copy, rearrange, and market a competing database at a

fraction of the cost incurred by the original database maker. COMM. ON ISSUES IN

THE

TRANSBORDER FLOW OF SCIENTIFIC DATA, U.S. NAT’L COMM. FOR CODATA,

COMM’N ON PHYSICAL SCIENCES, MATHEMATICS, AND APPLICATIONS, AND NAT’L

RESEARCH COUNCIL, BITS OF POWER: ISSUES IN GLOBAL ACCESS TO SCIENTIFIC DATA

1-52 (National Academy Press forthcoming 1997) [hereinafter G

LOBAL ACCESS].

Although database makers can erect technological fences to protect their invest-

ments (e.g., encryption algorithms), those fences suffer limitations which render

them less than optimal as a means of protection. See, e.g., Henry H. Perritt, Jr.,

Property and Innovation in the Global Information Infrastructure, 1996 U. CHI. LEGAL

F. 261, 288 (1996) (discussing market, technological, and legal limitations as tools

to protect intellectual property on the Internet). In addition, useable encryption

technology forces only the first purchaser to pay for a key to decrypt the encoded

information. Therefore, unless there is a fence, or other disincentive to prevent

the first purchaser from repackaging the database maker’s investment,

encryption suffers the same weakness as the two-party deal discussed infra in

Part I.

702 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

parasitic competitors, users, and information Samaritans,

5

unless the database maker renders the database less valu-

able, or more expensive,

6

to the information user by crea-

tively “selecting and arranging” enough disparate data to

erect the fence

7

afforded by copyright,

8

erects the two-party

5. Information Samaritans refers to parties who, for noneconomic reasons,

extract data from databases without paying the database maker and make it

freely available to the public. Cf. United States v. LaMacchia, 871 F. Supp. 535,

536 (D. Mass. 1994). In LaMacchia, a federal grand jury indicted LaMacchia, a 21-

year-old M.I.T. student, for devising a scheme in violation of 18 U.S.C.A. § 1343

(West Supp. 1996), the wire fraud statute, to defraud software manufacturers and

vendors. Id. LaMacchia encouraged users to upload popular software and

games to an electronic bulletin he named Cynosure, and transferred the pro-

grams to a second location where users with access to the Cynosure password

could freely download the programs. Id. Because LaMacchia had apparently not

acted for the purposes of securing “commercial advantage or private financial

gain,” he could not be prosecuted under the criminal copyright statute, 17

U.S.C.A. § 506 (West Supp. 1996). Id. at 545; see also Copyright Law of the United

States of America, Pub. L. No. 94-553, 90 Stat. 2541 (1976) (current version at 17

U.S.C.A. §§ 101-810 (West Supp. 1996)). Citing, among other factors, the failure

to allege violation of the criminal copyright statute, id. at 542-43, and relying

heavily on the reasoning of Dowling v. United States, 473 U.S. 207 (1985), the dis-

trict court granted LaMacchia’s motion to dismiss, but nonetheless noted that if

the indictment was true, LaMacchia’s behavior was “at best heedlessly irrespon-

sible, and at worst . . . nihilistic, self-indulgent, and lacking in any fundamental

sense of values.” LaMacchia, 871 F. Supp. at 545. This Article does not discuss

the conflicting policies presented by criminal prosecution of information Samari-

tans. See Teddy C. Kim, Taming the Electronic Frontier: Software Copyright Protec-

tion in the Wake of United States v. LaMacchia, 80 MINN. L. REV. 1255 (1996) (argu-

ing that criminal prosecution of noncommercial copyright infringers is

inappropriate). However, this Article does proceed on the assumption that ex-

tracting data from databases and making it freely available, regardless of the mo-

tive for such behavior, destroys incentives to invest in database creation. See

GLOBAL ACCESS, supra note 4, at 8.

6. For example, to secure copyright protection, many database makers inject

“value-added” data, which increases the price of the database to the information

user without any corresponding increase in utility. See Jane C. Ginsburg, No

“Sweat”? Copyright & Other Protection Works of Information, 92 COLUM. L. REV. 338,

344-47 (1992). The additional data adds “value” only in the sense that it enables

the database maker to secure some degree of copyright protection. Id. at 347.

7. Cf. Wendy J. Gordon, Asymmetric Market Failure and Prisoner’s Dilemma in

Intellectual Property, 17 U. DAYTON L. REV. 853, 855 (1992) (noting that the right to

exclude afforded authors functions much in the same way as fences do for real

property owners).

8. Copyright generally protects only the creative selection and arrangement

of factual information. See Berne Convention for the Protection of Literary and

1997] EUROPEAN DATABASE DIRECTIVE 703

wall afforded by contract,

9

or is otherwise able to secure the

haphazard protection whimsically afforded by unfair com-

petition law.

10

Artistic Works of September 9, 1886, completed at Paris on May 4, 1896, revised

at Berlin on Nov. 13, 1908, completed at Berne on Mar. 20, 1914, revised at Rome

on June 2, 1928, revised at Brussels on June 26, 1948, and revised at Stockholm on

July 14, 1967 (with Protocol regarding developing countries), completed at

Stockholm on July 14, 1967, art. 2(5), 828 U.N.T.S. 221 [hereinafter Berne Conven-

tion] (protecting collections of works, which by reason of selection and arrange-

ment of their contents, constitute intellectual creations); 17 U.S.C.A. §§ 101, 103

(West Supp. 1996) (protecting compilations of data that are selected, coordinated,

or arranged in such a way that the compilation constitutes an original work of

authorship). With regard to copyright protection, the Database Directive is in

accord. See Database Directive, supra note 1, recitals 15-16, O.J. L 77/20, at 21,

art. 3(1), O.J. L 77/20, at 25 (1996) (databases, which by reason of selection or ar-

rangement of their contents constitute intellectual creations, shall be protected by

copyright); Diplomatic Conference on Certain Copyright and Neighboring

Rights Questions, Draft Treaty on New Copyright Developments, art. 5 WIPO

Doc. CRNR/DC/89 (Dec. 20, 1996) (compilations of data, which by reason of se-

lection or arrangement of their contents, constitute intellectual creations are pro-

tected), adopted by Diplomatic Conference on Certain Copyright and Neighboring

Rights Questions, Geneva, Dec. 20, 1996; see also J. H. Reichman, Universal Mini-

mum Standards of Intellectual Property Under the TRIPs Component of the WTO

Agreement, 29 INT’L LAW. 345, 373 (1995) (noting that the Agreement on Trade-

Related Aspects of Intellectual Property Rights (“TRIPS”) fails to offer the protec-

tion provided by the Database Directive’s sui generis regime). But see Copyright,

Designs and Patents Act, 1988, 2 Eliz. 2, ch. 48, §§ 2-3 (Eng.) (protecting “sweat of

the brow” databases in the United Kingdom); Reichman, supra note 3, at 804 n.33

(citing W. R. CORNISH, INTELLECTUAL PROPERTY: PATENTS, COPYRIGHT,

TRADEMARKS AND ALLIED RIGHTS 268-69 (2d ed. 1989) for the United Kingdom’s

practice of extending copyright to small change literary productions provided

they exhibit “skill, judgment, and labor”); see also Ginsburg, supra note 6, at 372

n.161 (noting that the Nordic Catalogue Rule provides protection to unoriginal

compilations for 10 years). The Database Directive permits derogation from its

terms so that databases protected under less-exacting EC copyright regimes on

the Database Directive’s effective date will continue to enjoy that protection until

their term expires. Database Directive, supra note 1, art. 14(2), O.J. L 77/20, at 27

(1996); cf. WIPO Proposal, supra note 1, art. 7(3) WIPO Doc. CRNR/DC/6 (Aug.

30, 1996) (rights granted by treaty are independent of any protection afforded to

the database maker by national legislation). Similarly, the sui generis rights

granted by the three regimes are without prejudice to any other rights. See Data-

base Directive, supra note 1, art. 13, O.J. L 77/20, at 27 (1996); H.R. REP. NO. 3531,

supra note 1, § 9 (c); WIPO Proposal, supra note 1, art. 10(2) WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996).

9. See generally

JOHN D. CALAMARI & JOSEPH M. PERILLO, CONTRACTS 4, 9 (3d

ed. 1987).

10. J. H. Reichman, Legal Hybrids Between the Patent and Copyright Paradigms,

704 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Recognizing the need for a “stable and uniform legal pro-

tection regime” to stimulate database creation in the Euro-

pean Community (“EC”),

11

the EC struck out boldly by erect-

ing a toll fence

12

to protect the investment of database

94 COLUM. L. REV. 2432, 2476 (1994) (noting that the development of unfair com-

petition proceeds on a “hit-or-miss basis that varies with the outlook of single

judges . . . [without any] well-defined objects of protection, . . . standards of eli-

gibility, and [with] few safeguards to protect the interests of all concerned”).

11. Database Directive, supra note 1, recital 12, O.J. L 77/20, at 20 (1996).

The United States’ share of the database market is approximately twice the EC’s

market share. See Explanatory Statement at COM(92)24 final-SYN 393 accompa-

nying the Council Directive Initial Proposal, art. 8, O.J. C 156/03, at 9 (1992)

[hereinafter Initial Proposal]; see also Debra B. Rosler, The European Union’s Pro-

posed Directive for the Legal Protection of Databases: A New Threat to the Free Flow of

Information, 10 HIGH TECH. L.J. 105, 109-10 (1995) (explaining that Great Britain

garners approximately 65% of the revenues from the EC database market). Is it

merely a coincidence that the two countries with the biggest database market

share, the United States and Great Britain, have historically protected “noncrea-

tive” labors of database makers through “sweat of the brow” and “industrious

collection” rationales? See Pamela Samuelson, Missing Foundations of The Pro-

posed European Database Directive (1994) (on file with Professor Jerome H.

Reichman, Vanderbilt University School of Law). For a comparison of the two

rationales, compare Jane C. Ginsburg, Creation and Commercial Value: Copyright

Protection for Works of Information, 90 COLUM. L. REV. 1865 (1990) (discussing

United States “sweat of the brow” rationale) with Colin Tapper, An Aspect of

Copyright in Data Bases, 14 N. KY. L. REV. 169 (1987) (explaining Great Britain’s

“industrious collection” rationale). But see Mark Powell, The European Union’s

Database Directive: An International Antidote to the Side-Effects of Feist?, in 3

FORDHAM INTERNATIONAL INTELLECTUAL PROPERTY LAW & POLICY (Hugh C. Han-

sen ed., forthcoming 1997) (citing the European Parliament’s Committee on Eco-

nomic and Monetary Affairs opinion that “the economic situation in the database

sector is such that urgent actions are not as yet required”).

12. Cf. Gordon, supra note 7, at 855. Gordon argues that groups seeking new

fences “should be prepared to show that their current fences are inadequate to

provide adequate incentives.” Id. To justify their need for a new fence, database

makers need only point to Feist Publications, Inc. v. Rural Telephone Service Co., 499

U.S. 340 (1991), and its implications. See Ginsburg, supra note 6, at 343

(“Feist . . . calls into question the ability of copyright . . . to secure meaningful

coverage to those compilations that do meet the initial test of copyrightablity.”).

In Feist, the United States Supreme Court cloaked free riders who appropriate

the “noncreative”

labors of data collectors with constitutional protection by en-

dorsing the thin copyright doctrine. Feist, 499 U.S. at 357-59. Apparently, only

the creative selection and arrangement of facts deserve copyright protection.

Unfortunately, for many database makers, the exercise of creativity in the selec-

tion and arrangement of data is constrained by societal expectations, notions of

functionality, and “user-friendliness” considerations. See Ginsburg, supra note 6,

at 343; Robert C. Denicola, Copyright in Collections of Facts: A Theory for the Protec-

1997] EUROPEAN DATABASE DIRECTIVE 705

makers.

13

Emboldened by, or perhaps scurrying to catch up

with,

14

the vanguard approach

15

embodied in the EC’s Data-

base Directive,

16

the United States and the World Intellectual

Property Organization (“WIPO”) have proposed similar pro-

tection for these valuable business tools.

17

Like the Database

Directive, the United States and WIPO proposals (“Propos-

als”), respectively known as the Database Investment and In-

tellectual Property and Antipiracy Act of 1996 (“H.R. 3531”

or “U.S. Proposal”)

18

and the Draft Treaty on Intellectual

tion of Nonfiction Literary Works, 81 COLO. L. REV. 516, 527-30 (1981).

13. The Database Directive extends copyright protection to “authors” of da-

tabases, Database Directive, supra note 1, arts. 3-5, O.J. L 77/20, at 25 (1996), but

extends sui generis protection to a broader group: database makers. Id. art. 7(1),

O.J. L 77/20, at 25 (1996). The Database Directive defines database makers as

those who take the “initiative and risk of investing.” Id. recital 41, O.J. L 77/20,

at 23 (1996). Unlike the Database Directive, which leaves ownership of em-

ployee-made databases up to national legislation, see id. recital 29, O.J. L 77/20,

at 22 (1996), both the U.S. and WIPO Proposals define database maker in such a

way as to preclude employees and subcontractors from gaining ownership

rights. See H.R. REP. NO. 3531, supra note 1, § 2; WIPO Proposal, supra note 1, art.

2(3) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996) (database maker is the person with

control and responsibility for the undertaking of a substantial investment). This

Article does not address the copyright protection afforded database makers by

the Database Directive. Id. art. 7(1), O.J. L 77/20, at 25 (1996).

14. See Paul Waterschoot, An Overview of Recent Developments in Intellectual

Property in the European Union, in 3 FORDHAM INTERNATIONAL INTELLECTUAL

PROPERTY LAW & POLICY, supra note 11 (explaining that the Database Directive

“has placed the Community far ahead of its partners”); 142 CONG. REC. E890-04

(daily ed. May 23, 1996) (statement of Sen. Moorhead) (explaining that the “Di-

rective could place U.S. firms at an enormous competitive disadvantage

throughout the entire European market”).

15. See Powell, supra note 11, at 50 n.2 (noting that the EC’s “vanguard ap-

proach to the legal protection of databases” parallels the United States’ adoption

of a sui generis regime, the Semiconductor Chip Protection Act of 1984, Pub. L.

No. 98-620, 98 Stat. 3347 (1984) (codified as amended at § 17 U.S.C.A. §§ 901-914

(West Supp. 1996)), to protect semiconductors).

16. Council Directive No. 96/9, O.J. L 77/20 (1996). The Database Directive

can be logically divided into three sections: Articles 3-6 govern the copy-

rightability of databases; Articles 7-11 govern the sui generis extraction and reuti-

lization rights; and Articles 1, 2, and 12-16 are common provisions applicable to

both bundles of rights. See generally Council Directive No. 96/9, O.J. L 77/20

(1996). This Article does not address the copyright provisions of the Database

Directive.

17. See supra note 1-2.

18. H.R. REP. NO. 3531, 104th Cong., 2d Sess. (1996).

706 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Property in Respect of Databases (“WIPO Proposal”),

19

would create sui generis regimes

20

protecting investment, as

such, in databases.

Some commentators argue that the Database Directive,

and the Proposals it has spawned, are arguably the “most

deviant” examples of the trend toward sui generis protection

for intellectual property falling between the cracks of the

mature paradigms of patent and copyright.

21

These com-

mentators also argue that the Database Directive is “one of

the least balanced and most potentially anti-competitive in-

tellectual property [regimes] ever created,”

22

and join others

who argue that the Database Directive will stifle access to in-

formation, retard competition in the database industry, and

impede basic scientific research.

23

This Article argues that the Database Directive’s sui

generis regime, at least with respect to publicly or privately-

19. Diplomatic Conference on Certain Copyright and Neighboring Rights

Questions, Draft Treaty on Intellectual Property in Respect of Databases, WIPO

Doc. CRNR/DC/6 (Aug. 30, 1996).

20. Intellectual property laws protecting subject matter outside the scope of

the mature intellectual property paradigms of patent and copyright law are

commonly deemed “sui generis.” See Reichman & Samuelson, supra note 4, at 6

n.6 (citations omitted). Sui generis literally means “of its own class.” See BLACK’S

LAW DICTIONARY 1434 (6th ed. 1990).

21. See Reichman & Samuelson, supra note 4, at 53-54; see generally

Reichman, supra note 10, at 2453-99 (discussing legal regimes protecting intellec-

tual property falling between the cracks of the mature patent and copyright

paradigms); Reichman, supra note 4, at 517 (suggesting the need for a new intel-

lectual property paradigm, not based upon exclusive property rights, that “looks

beyond art and inventions . . . [and] deal[s] directly with the pervasive threat of

market failure facing investors in unpatentable, noncopyrightable innovation”).

22. Reichman & Samuelson, supra note 4, at 79.

23. See Rosler, supra note 11, at 146; Charles Von Simson, Feist or Famine:

American Database Copyright as an Economic Model for the European Union, 20

B

ROOK. J. INT’L L. 729, 766-68 (1995); Letter from Bruce Alberts et al., President,

National Academy of Sciences, to Michael Kantor, Secretary of Commerce (Oct.

9, 1996) (on file with author) [hereinafter “Academy Letter”]; Reichman &

Samuelson, supra note 4, at 55; GLOBAL ACCESS, supra note 4, at 35. “Basic re-

search” seeks to further scientific knowledge for its own sake, without regard to

specific applications. See N

ATIONAL SCIENCE BOARD, SCIENCE & ENGINEERING

INDICATORS 4-9 (15th ed. 1996).

1997] EUROPEAN DATABASE DIRECTIVE 707

available data (“nonproprietary data”),

24

serves some of the

same procompetitive functions as does the law of trade se-

crets

25

without requiring secrecy.

26

While this Article recog-

nizes that the Database Directive’s sui generis regime could

result in de facto “fact monopolies,” or otherwise enable da-

tabase makers to charge monopoly rents for sole-source or

proprietary data,

27

this Article argues that ameliorative

measures available in the EC are sufficient to enable compe-

tition, the ultimate guardian of consumers in market econo-

mies,

28

to perform its protective role in the EC, even in the

case of sole-source data. However, the lack of an interna-

tional competition law or other international sword,

29

which

24. Nonproprietary data are data that can be independently generated, col-

lected, or obtained from more than one source. Compare Initial Proposal, supra

note 11, art. 8, O.J. C 156/03, at 9 (1992) with Council Directive Amended Pro-

posal, art. 11, O.J. C 308/01, at 13-14 (1993) [hereinafter Amended Proposal] (de-

fining “sole-source data” as data that cannot be “independently created, collected

or obtained from any other source”).

25. See ProCD v. Zeidenberg, 86 F.3d 1447, 1454-55 (7th Cir. 1996) (citing

Rockwell Graphic Sys. v. DEV Indus., 925 F.2d 174, 180 (7th Cir. 1991)); cf.

Reichman, supra note 10, at 2439-40.

26. Under the Database Directive, second comers seeking to extract data

from a protected database must either obtain a license from the database maker

to extract the data, or independently generate the data. See infra notes 244-54 and

accompanying text; cf. Reichman, supra note 10, at 2438-40 (noting that trade se-

cret laws force second comers to license unpatentable know-how or to obtain the

know-how by proper means); Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 490

(1974) (noting that trade secret law does not forbid independent creation, reverse

engineering, or other fair and honest means).

27. See supra note 24 (defining “nonproprietary data”).

28. As the Supreme Court has explained: “Basic to the faith that a free

economy best promotes the public wealth is that goods must stand the cold test

of competition; that the public, acting through the market’s impersonal judg-

ment, shall allocate the Nation’s resources and thus direct the course its eco-

nomic development will take . . . .” Jefferson Parrish Hosp. v. Hyde, 466 U.S. 2,

12 (1984) (quoting Time-Picayune Publishing Co. v. United States, 345 U.S. 594,

605 (1953)); see also Digital Equip. v. Uniq Digital Technologies, 73 F.3d 756, 762

(7th Cir. 1996) (“Competition among manufacturers fully protects buyers who

accurately calculate life-cycle costs.”); cf. Rockwell Graphic Sys., 925 F.2d. at 180

(discussing importance of trade secrets to preserve competition in intellectual

property); Zeidenberg, 86 F.3d at 1453 (“Competition among vendors . . . is how

consumers are protected in a market economy.”).

29. See Eleanor M. Fox, Trade, Competition, and Intellectual Property—TRIPS

708 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

would enable fair followers

30

to compete directly, or indi-

rectly with sole-source database makers in value-added

markets, suggests a need to build a limitation into any inter-

national sui generis regime protecting investment in data-

bases.

31

This Article discusses and analyzes the EC Database Di-

rective’s sui generis regime,

32

which, unlike copyright law,

casts its protection based upon economic principles, rather

than broad cultural policies.

33

Part I examines briefly the

general characteristics of the database markets and the pro-

tection afforded databases by existing legal regimes. Part II

compares the substantive provisions of the Database Direc-

tive and the Proposals. Part III analyzes the Database Direc-

tive’s sui generis regime, using the factual background of

ProCD, Inc. v. Zeidenberg,

34

a case whose facts demonstrate

the vulnerability of databases to parasitic behavior, and con-

cludes that the Directive adequately balances the public’s

need for information access with the need for production in-

centives within the EC.

and its Antitrust Counterparts, 29 VAND. J. TRANSNAT’L L. 481, 485-91 (1996) (dis-

cussing the significant differences in the world’s two prominent competition law

systems: European community competition law and United States antitrust

law). For an analysis of the interaction between EC competition law and intellec-

tual property law, see Jan Corbet, The Law of the EEC and Intellectual Property, 13

J.L. & COM. 327 (1994).

30. Cf. J. R. Reichman, From Free Riders to Fair Followers: Global Competition

Under the TRIPS Agreement, N.Y.U.

J. INT’L L. & POL. (forthcoming 1997).

31. See infra note 35 and accompanying text.

32. See supra notes 1, 16.

33. See, e.g., Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417,

429 (1984) (explaining that copyright law makes economic reward secondary to

the higher public purpose of providing the public access to the author’s creative

genius) (citations omitted); see also Reichman, supra note 10, at 2452-53 (noting

that copyright promotes cultural policies at the expense of efficient allocation of

resources); see also Pamela Samuelson, Comments on Gerald Dworkin’s Article on

Copyright, Patent or Protection for Computer Programs, in 1 F

ORDHAM INTERNA-

TIONAL

INTELLECTUAL PROPERTY LAW & POLICY 183 (Hugh C. Hansen ed., 1996)

(noting that the Database Directive is based upon “competition policy, rather

than on the . . . Romantic concepts of authorship embedded in . . . Continental

European [copyright]”).

34. 908 F. Supp. 640 (W.D. Wis. 1996), rev’d, 86 F.3d 1447 (7th Cir. 1996).

1997] EUROPEAN DATABASE DIRECTIVE 709

Part IV argues that the Database Directive’s sui generis

regime would, if modified by incorporating a compulsory li-

cense

35

loosely modeled on the compulsory license United

States copyright mandates for musical works,

36

provide a

sufficiently pro-competitive international model to protect

the public’s interest in competition in the information mar-

ket. So modified, the Database Directive’s sui generis regime

would permit fair followers to follow by honest means,

while protecting the “sweat of the brow” intellectual efforts

37

and investments of database makers. Part IV then discusses

the risks and opportunities to researchers who currently op-

erate in noncompetitive, information-subsidized environ-

ments. Finally, this Article concludes by suggesting that a

sui generis regime protecting investment in databases could

35. See Initial Proposal, supra note 11, art. 8, O.J. C 156/03, at 9 (1992);

Amended Proposal, supra note 24, art. 11, O.J. C 308/01, at 13-14 (1993) (mandat-

ing compulsory licenses on “fair and nondiscriminatory terms” in the case of

sole-source data); cf. Reichman, supra note 10, at 2539-44 (proposing off-the-rack

liability regime with built-in compulsory license); see also Pamela Samuelson et

al., A Manifesto Concerning the Legal Protection Of Computer Programs, 94 COLUM. L.

REV. 2308 (1994); cf. Reichman & Samuelson, supra note 4, at 81 (proposing short

period of pro-competitive lead-time followed by an automatic license).

36. See 17 U.S.C.A. § 115.

37. Notwithstanding Feist’s epistemologically erroneous incantation that

facts merely exist and await discovery, and therefore cannot be original, Feist,

499 U.S. at 347-48, the author’s contributions are no more original than the re-

searcher’s contributions: both translate and recombine data from the public do-

main. See Jessica Litman, The Public Domain, 39 E

MORY L.J. 965, 1016 (1990) (argu-

ing that facts do not exist until researchers “sift through available evidence,

design new avenues of inquiry, choose among a myriad of conflicting indicia,

and supply interpretive paradigms”); see also Wendy J. Gordon, Reality as Arti-

fact: From Feist to Fair Use, 55 LAW & CONTEMP. PROBS. 93 (1992); cf. Zechariah

Chafee, Jr., Reflections on the Law of Copyright, 45 COLUM. L. REV. 503, 511 (1945)

(explaining that authors stand on the shoulders of the cultural giant). The ana-

lytical soundness of this conceptual truism is demonstrated by the narrow scope

afforded Feist. See, e.g., CCC Info. Servs., Inc. v. Maclean Hunter Mkt. Reports,

Inc., 44 F.3d 61, 67 (2d Cir. 1994) (holding that the selection and arrangement of

used car valuations was creative), cert. denied, 116 S. Ct. 72 (1995); Key Publica-

tions v. Chinatown Today Publishing Enters., 945 F.2d 509 (2d Cir. 1991) (hold-

ing that the selection of businesses to be included in a telephone directory was

creative); Kregos v. Associated Press, 937 F.2d 700 (2d Cir. 1991) (holding that

the selection of categories in creating a baseball pitching form was creative).

710 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

effect a subtle, and perhaps favorable, shift in societal re-

sources from the pursuit of entertainment to the pursuit of

knowledge.

I. DATABASE PROTECTION UNDER EXISTING LEGAL REGIMES

Although most databases find some degree of protection

under the veil of copyright law, between the two-party wall

afforded by contract, or within the whimsical embrace of un-

fair competition law, many commentators agree that these

legal regimes, both singularly and collectively, fail to pro-

vide protection adequate to ensure the certainty of return on

investment required to enhance and stimulate worldwide

production of databases.

38

To reveal the gap-filling func-

38. See, e.g., Statement on the NII Copyright Protection Act of 1995: Hearings

Before the Subcomm. on Courts and Intellectual Property House Comm. on the Judici-

ary, 105th Cong., 2d Sess. (1997) (statement of Barbara A. Munder, Senior Vice

President, McGraw-Hill Companies, testifying on behalf of the Information In-

dustry Association “IIA”) (“Information is literally and figuratively at the heart

of the National Information Infrastructure. . . .”); id. (“[W]ithout effective protec-

tion, [producers of information content] cannot risk . . . investment in cyberspace

where it is so easy to copy, retransmit and alter our property without our per-

mission, and often without our knowledge.”). The Information Industry Asso-

ciation is a trade association of 550 companies that provides information prod-

ucts and services on a worldwide basis. See Bradford L. Smith, Creating the Global

Information Society: Looking Ahead, in 2 FORDHAM INTERNATIONAL INTELLECTUAL

PROPERTY LAW & POLICY (Hugh C. Hansen ed., forthcoming 1997); Reichman &

Samuelson, supra note 4, at 7; Denicola, supra note 12, at 528-30; Ginsburg, supra

note 6, at 340-42; Dennis S. Karjala, Copyright in Electronic Maps, 35 J

URIMETRICS J.

395, 396 (1995) (arguing that a new approach, which may require either amend-

ing the Copyright Act or adopting a sui generis database protection statute, is

needed to protect electronic maps); see also Database Directive, supra note 1, re-

cital 12, O.J. L 77/20, at 20 (1996); WIPO Proposal, supra note 1, Note 1.05 WIPO

Doc. CRNR/DC/6 (Aug. 30, 1996) (“[C]ontinued investment . . . in the develop-

ment and refinement of databases . . . will not take place unless a stable and uni-

form regime of legal protection is established . . . .”); cf. 142 CONG. REC. E890-04

(daily ed. May 23, 1996) (statement of Sen. Moorhead) (introducing H.R. 3531 to

“encourage continued investment in the production and distribution of valuable

new databases”); cf. Rosler, supra note 11, at 107 (the inherent vulnerability of

electronic databases amplifies the need for economic protection while preserving

the “free flow of information”); Dennis S. Karjala, Misappropriation as a Third Intel-

lectual Property Paradigm, 94

COLUM. L. REV. 2594, 2595 (1994) [hereinafter Karjala

II] (suggesting a new approach to prohibit methods of acquiring information

1997] EUROPEAN DATABASE DIRECTIVE 711

tions that could be served by a sui generis regime designed to

protect the investments of database makers, this part briefly

examines various “database markets.” This part then ex-

plores the relative strengths and weaknesses of existing legal

regimes that afford some protection to these valuable busi-

ness tools.

A. General Characteristics of the Splintered Database Markets

Databases, and their respective commercial markets, vary

as much as the needs and capabilities of the information

consumers patronizing the database market.

39

Although the

commercial “database market” is difficult to define, the

competitive battle for the information user’s patronage ap-

pears to be waged in at least three somewhat distinct, but

overlapping, markets: the “one-stop-shopping” market; the

“problem-focused” market; and the “industry-focused” mar-

ket.

40

which, if permitted, would result in disincentives to create desirable works). The

failure of existing legal regimes to adequately protect database makers can be

partially attributed to the peculiar characteristics of information: “a commod-

ity . . . particularly embarrassing for the achievement of optimal allocation.” See

W. KIP VISCUSI ET AL., ECONOMICS OF REGULATION AND ANTITRUST 831-32 (2d ed.

1995) (quoting Kenneth J. Arrow, Economic Welfare and the Allocation of Resources

to Invention, in THE RATE AND DIRECTION OF INVENTIVE ACTIVITY 609, 620 (National

Bureau of Economic Research ed., 1962)). That is, optimal information utilization

occurs when information is free, while optimal information production occurs

only when producers expect to appropriate the economic value of their invest-

ments. Id.

39. Databases range in size and complexity from the Internet, which is con-

ceptually a database, to the Medicare and Medicaid Guide (CCH), a compilation

of general and specific information relating to health care law, available in paper

and electronic form, to simple tables of information.

40. Lloyd A. Fletcher, Searcher, I

NFO. TODAY, Sept. 1, 1995, available in West-

law, Allnewsplus File; Telephone Interview with Anne Griffith, Senior Research

Analyst at the Software Publishers Association, Washington, D.C. (Nov. 8, 1996)

[hereinafter “Griffith Interview”] (explaining that the database market consists of

many niche markets). Examples of databases serving the broader market include

NEXIS (current and archived news and information) and ProQuest Direct (jour-

nals, newspapers, and dissertation abstracts). FRED NEVIN, BUSINESS

INFORMATION MARKETS 1999: THE STRATEGIC OUTLOOK 67-77 (Lorraine Sileo et al.

eds., 1996 C

OWLES/SIMBA INFORMATION). Business databases generally fall in

seven topical areas (listed in descending market share order): financial (real-

712 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Database vendors in the “one-stop-shopping” market of-

fer general information content to a broad customer base.

41

In contrast, database vendors in the “problem-focused” mar-

ket offer specific information content, focused on particular

problems, to industry-wide groups,

42

while database ven-

dors in the “industry-focused” market offer both general and

specific information content to specific industry and profes-

sional markets such as medical, legal, government, public re-

lations, and news.

43

Generally speaking, database vendors offering general

information content serve a broad market and incur lower

data collection costs, which permits those database vendors

to charge lower fees while remaining competitive.

44

Con-

versely, because database vendors offering highly special-

ized information content generally serve a narrower market

and incur higher data collection costs, those database makers

typically charge higher fees to earn a return on their invest-

ment.

45

As a result, information users seeking specialized in-

time and historical), marketing, credit, research, market-specific, le-

gal/tax/public records, and general and business news. Id.

41. See Fletcher, supra note 40 (noting that even within the broader market

each vendor has its own relative subject matter strengths).

42. Id. (citing LEXIS/NEXIS and Knight-Ridder’s Business-Base as exam-

ples); see also NEVIN, supra note 40, at 72 (listing Telerate as a provider of a data-

base focusing on the price of government securities and bonds).

43. Fletcher, supra note 40 (citing LEXIS/NEXIS’ AnswerPaks and KRI’s

Science-Base as examples); see also N

EVIN, supra note 40, at 75-76 (listing Equifax

Insurance Information Services (insurance industry), T/SF Communications

(trucking industry), and ARI Network Services (agriculture and environmental)

as providers of market-specific databases).

44. Griffith Interview, supra note 40; see also Perritt, supra note 4, at

45. Griffith Interview, supra note 40 (noting other factors driving costs in

specialized markets such as increased collection costs because data are not as

widely available, and higher data maintenance costs associated with dynamic

data). the cost of originating intellectual prop-

erty can be expressed as:

co = cc + cc + cm + cr where cc is the cost of creation, e.g. payments to

the author; cc is the cost of . . . preparing the information for publica-

tion; cm is the cost of marketing, including promotional expenses, dis-

tribution costs and costs of billing and collecting; and cr is the cost of

copying (reproduction).

1997] EUROPEAN DATABASE DIRECTIVE 713

formation will generally pay a higher price than information

users seeking general information.

46

Information users also generally find pricing structure

variations within each market segment.

47

For example, sub-

scription models may charge a basic fee for unlimited ac-

cess,

48

or a basic fee for a fixed level of access plus an addi-

tional time-based fee for access beyond the fixed level.

49

Transactional pricing models may charge based upon the

number of searches performed or upon the volume of in-

formation retrieved or extracted.

50

Other pricing models al-

low free public access to a certain amount of information,

and restrict access to more specialized information to paid

subscribers.

51

Finally, some pricing models incorporate fea-

tures of both the subscription and transaction models

52

and

permit the information user to choose the pricing scheme

that maximizes usability while minimizing costs.

53

Perritt, supra note 4, at

46. Griffith Interview, supra note 40; Perritt, supra note 4, at 276

47. See Perritt, supra note 4, at (describing various pricing options avail-

able through Westlaw).

48. For example, CD-ROM subscriptions typically provide unlimited access

to an optical disk, along with periodic updates, for a fixed fee; CCH CD-ROM

and Online for Windows (CCH) offers the Medicare and Medicaid Guide along

with various other health law materials on optical disk with monthly updates;

and local Internet access providers, as well as some centralized dial-up services

such as America Online, offer an option for purchasing unlimited access.

49. Griffith Interview, supra note 40. For example, the CCH product noted

above offers access to an accompanying online service, charged on an hourly ba-

sis, that provides access to the latest health and tax law materials. See supra note

48.

50. Id.

51. For example, Environmental RouteNet, a service offered by Cambridge

Scientific, focuses on environmental information. Routenet also engages in price

discrimination: individual users may access for $50 per month, while institu-

tional prices start at $8,000 per year for eight users. See Paul Blake, Database Tra-

ditionals Get Caught Up in the Web, I

NFO. TODAY, Sept. 1, 1995, available in Westlaw,

Allnewsplus File; see also United States Courts Southern District of Texas (visited

May 6, 1997) <http://www.txs.uscourts.gov> (permitting free access to certain

data via the Internet but imposing time-based fees via a stand alone BBS

(PACER) for other data).

52. Examples include LEXIS/NEXIS and Westlaw.

53. Id.; see also John B. Kennedy & Shoshana R. Dweck, Publishers, Authors

714 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Although variations in the database market structure and

price models may reflect some degree of competition,

54

commentators have noted that the market operates at a

suboptimal level because of the failure of existing legal re-

gimes to protect adequately the investment required to pro-

duce valuable, commercial databases.

55

B. Protection Afforded Database Makers by Existing Legal

Regimes

Database makers and rightholders

56

seeking to market

their databases can obtain some degree of protection from

existing legal regimes.

57

This section briefly explores, in

Battle Over Electronic Rights, NAT’L L.J., Oct. 28, 1996, at C17 (noting that publish-

ers are experimenting with tiered subscriptions, per-article, and time-based pric-

ing mechanisms).

54. But see Reichman & Samuelson, supra note 4, at 69 (noting that anecdotal

evidence suggests that database markets are “almost universally characterized

by a distinct absence of competition”). While this proposition may be true of

markets dominated by database makers who are also the sole source for the data

sought, the pricing options available to customers seeking the same type of in-

formation, see supra notes 47-53 and accompanying text, the high level of merger

and acquisition activity in the business database market, and falling operating

margins, indicate some level of competition in the database market. See NEVIN,

supra note 40, at 81.

55. See supra notes 5, 8; Reichman & Samuelson, supra note 4, at 55 (noting

that the risk of market failure tends to keep the production of information goods

at suboptimal levels); see also Reichman, supra note 10, at 2491 (noting that copy-

right’s creativity requirement could “exclude many of the most commercially

and scientifically important databases”).

56. Rights under the three sui generis regimes initially vest in the database

maker, but are freely transferable thereafter. See Database Directive, supra note 1,

art. 7(3), O.J. L 77/20, at 26 (1996); WIPO Proposal, supra note 1, art. 4(2) WIPO

Doc. CRNR/DC/6 (Aug. 30, 1996); H.R.

REP. NO. 3531, supra note 1, § 2 (defining

“database owner”).

57. Database makers not seeking to market their databases who also keep

their databases confidential might obtain trade secret protection. See, e.g., Mis-

sissippi Uniform Trade Secrets Act, MISS. CODE ANN. §§ 75-26-1 to -19 (1996);

Economic Espionage Act of 1996, Pub. L. No. 104-194, 110 Stat. 3488 (1996) (codi-

fied at 18 U.S.C.A. §§ 1831-1839 (West Supp. 1996)) (making theft of trade secrets

a federal crime); RESTATEMENT (THIRD) OF UNFAIR COMPETITION LAW §§ 39-45

(1995). Database makers may also benefit from criminal laws designed to deter

computer-related crime. See Rosler, supra note 11, at 131 (stating that such laws

have been enacted in 49 states).

1997] EUROPEAN DATABASE DIRECTIVE 715

turn, the strengths and weaknesses of copyright, contract,

and unfair competition law, three regimes that offer some

protection to databases published in the marketplace.

58

1. Protection Afforded by Copyright

One might expect that copyright, an institution designed

to “promote the Progress of Science and useful Arts,”

59

would embrace a form of intellectual property with the po-

tential to enhance access to, and increase the availability of,

information.

60

Indeed, databases whose final form exhibits

creative selection and arrangement

61

secure protection

“against the world.”

62

However, as courts and commenta-

tors have recognized, the protection afforded by copyright

law to databases, except in a few European countries,

63

con-

sists of little more than a thin veil that quickly unravels as

content becomes more factual and necessary.

64

Some commentators have noted that creatively selecting

and arranging a database may reduce data availability and

accessibility.

65

For example, creatively selecting data ex-

58. In addition, the so-called Nordic “catalogue rule” protects “catalogues,

tables, and similar compilations in which a large number of particulars have

been summarized, including databases, for ten years after first publication.” See

Reichman, supra note 10, at 2492-93 (quoting Gunnar Karnell, The Nordic Cata-

logue Rule, in PROTECTING WORKS OF FACT 67-68 (E. J. Dommering & P. B. Hugen-

holtz eds., 1991)) (internal quotes omitted).

59. U.S.

CONST., art. I, § 8, cl. 8.

60. Paradoxically, this very purpose is often cited as the basis for limiting

copyright in factual information. See Feist, 499 U.S. at 354 (“Throughout history,

copyright law has recognized a greater need to disseminate factual works . . . .”)

(citation omitted).

61. See supra note 8.

62. See Zeidenberg, 86 F.3d at 1454.

63. See supra note 8.

64. See Glynn S. Lunney, Jr., Reexamining Copyrights’ Incentive Access Para-

digm, 49 V

AND. L. REV. 483, 567-71 (1996) (discussing copyright’s questionable

premise that by limiting the scope of protection for necessary works, society will

have greater access to such works); see supra notes 5, 12.

65. See Ginsburg, supra note 6, at 347-49; Denicola, supra note 12, at 530; see

supra note 38.

716 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

cludes potentially desirable information,

66

while creatively

arranging data, assuming arrangement is not functionally-

dictated, may frustrate information users.

67

In addition, be-

cause many database makers add superfluous original,

value-added data in an attempt to avail themselves of the

cloak of copyright, the selection and arrangement criteria

may increase the cost of databases without providing corre-

sponding gains to database users.

68

Other commentators have recognized that because copy-

right rests in a large degree upon broad cultural policies, and

not upon general economic considerations, copyright is ill-

suited to protect functional business tools like databases that

compete in the general market.

69

That is, by linking the exis-

tence and degree of protection to the superficial final form of

the product, rather than to total production costs, copyright

fails to offer sufficient production incentives to encourage

production of these valuable business tools,

70

and effectively

excludes “many of the most commercially and scientifically

important databases.”

71

Copyright, then, leaves databases whose content is fac-

tual or necessary or is otherwise considered noncreative (es-

pecially comprehensive, electronic databases that defer

66. See Ginsburg, supra note 6, at 344-45 (noting that the selection criterion

discourages comprehensive data collections); see also Reichman, supra note 10, at

2491.

67. See Ginsburg, supra note 6, at 344-45. Consequently, electronic database

makers often forego imposition of selection and arrangement criteria in favor of

permitting users to dynamically impose their own selection and arrangement

criteria. Id.

68. Id. at 344-46.

69. Cf. Reichman, supra note 10, at 2452-53 (noting that copyright promotes

certain cultural policies at the expense of economic efficiency); Reichman, supra

note 3, at 806 n.45 (copyright deliberately subordinates the price-setting function

of the market to broader cultural policies); see Denicola, supra note 12, at 516 (not-

ing that “the very vocabulary of copyright law is ill-suited to analyze property

rights in works of nonfiction[,]” such as databases).

70. See Denicola

, supra note 12, at 530.

71. See Reichman, supra note 10, at 2491 (creativity requirement could ex-

clude many of the most commercially and scientifically important databases).

1997] EUROPEAN DATABASE DIRECTIVE 717

completely to the information consumer’s judgment regard-

ing data arrangement and selection)

72

exposed to parasitic

competitors and information Samaritans who extract and re-

compile the raw, unfiltered data to produce competing

products.

73

2. Protection Afforded by Contract

Through mutually beneficial exchanges, contracts permit

the private ordering of resources (risk-allocation) essential to

the efficient functioning of a market economy.

74

Database

makers can obtain differing degrees of contractual protec-

tion, depending on whether they make the database publicly

available or keep it under private control.

75

The limits of contractual protection for publicly-available

databases are apparent after considering that unlike a copy-

right, which is enforceable against the world,

76

a contract is

enforceable only against the contracting parties.

77

Thus,

72. See Ginsburg, supra note 6, at 345 (computerized databases are designed

to permit users to impose their own arrangement and selection criteria).

73. See Reichman, supra note 10, at 2490-91. Information Samaritans might

distribute the data and effectively destroy a large portion of the maker’s market,

while parasitic competitors (“pirates”) might compete head-to-head with the da-

tabase maker, or might produce and market a value-added product before the

maker produces the product. Reichman & Samuelson, supra note 4, at 67-68. In

the latter case, the pirate might capitalize on advantages otherwise held by the

maker; e.g., established database makers have production facilities, managerial

experience, distribution channels, and reputational advantages. See Perritt, supra

note 4, at 279 (noting some of the market-related phenomena which tend to re-

duce the pirate’s ability to compete head-to-head with the maker).

74. See Zeidenberg, 86 F.3d at 1455 (citing American Airlines v. Wolens, 513

U.S. 219

, 115 S. Ct. 817, 824-26 (1995); see generally 3 E. ALLAN FARNSWORTH,

FARNSWORTH ON CONTRACTS § 12.3, at 153-54 n.1 (2d ed. 1990).

75. See G

LOBAL ACCESS, supra note 4, at 7-8 (discussing relative vulnerabili-

ties of public and private databases). As used herein, “publicly-available” means

that the database or its contents are physically available to the public in paper or

electronic form.

76. See Zeidenberg, 86 F.3d at 1454.

77. See generally J

OHN D. CALAMARI & JOSEPH M. PERILLO, CONTRACTS (3d ed.

1987). Although today’s technology permits preapproved computer-to-

computer transactions with no human interaction, Stewart I. Edelstein, Litigating

in Cyberspace: Contracts on the Internet, T

RIAL, Oct. 1996, at 16, “effective contract

718 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

unless the database maker can restrict redistribution, or con-

trol access to the database contents through technological

mechanisms, publicly-available databases are susceptible to

extraction and redistribution by parasitic competitors and in-

formation Samaritans who are not parties to the contract.

78

On the other hand, privately-controlled databases—those

databases to which access is restricted

79

—enable database

makers to “track and charge for every instance of electronic

access.”

80

Although such a database maker can theoretically

“reject the state-imposed cultural bargain”

81

of copyright law

and “impose monopoly prices and potentially oppressive

terms on users,”

82

there are several practical limits on her

ability to do so.

First, unless the database consists of sole-source data,

83

which in some cases could trigger antitrust scrutiny,

84

the

database maker faces the threat of market competition. Sec-

ond, database makers who make their product available

only through telecommunications devices necessarily limit

their market to those with access to such devices.

85

protection for content . . . [providers] is . . . problematic” on open (public) sys-

tems like the World Wide Web because “the transaction costs of effective nego-

tiations are high.” See Perritt, supra note 4, at 290-91.

78. See Reichman & Samuelson, supra note 4, at 65-68 (discussing the vul-

nerability of publicly-available databases); see also supra note 5 (discussing and

defining “information Samaritan”).

79. For example, access could be restricted by limiting distribution to con-

tracting parties via an electronic bulletin board or by other technological means

such as encryption devices. See supra note 4

80. Reichman & Samuelson, supra note 4, at 70.

81. Id.

82. Id. Note that the Database Directive voids contracts that violate certain

user rights. See Database Directive, supra note 1, art. 15, O.J. L 77/20, at 27

(1996); see also infra notes 167-70 and accompanying text.

83. See supra note 24.

84. See, e.g., John R. Wilke, FTC Charges ADP Formed a Monopoly in the Junk

Business, W

ALL ST. J., Nov. 15, 1996, at A4 (reporting that the Federal Trade

Commission filed suit against ADP for allegedly forming an information mo-

nopoly); see infra notes 136, 201.

85. For example, although the Internet operates in over 75 countries, it

reaches only approximately 25 million people who access the Internet via ap-

proximately two million personal computers. See Charles Clark, The Copyright

1997] EUROPEAN DATABASE DIRECTIVE 719

Third, contracts provide protection to databases, whether

publicly-available or privately-controlled, only if the parties

voluntarily abide by the contractual terms, or if the contrac-

tual terms are otherwise enforced by the judicial machin-

ery.

86

Commercially-minded courts have recognized that

enforcing private agreements is fundamental to market effi-

ciency and stability,

87

and that the enforcement of private

agreements respecting intellectual property serves the “same

procompetitive functions as does the law of trade secrets.”

88

However, some courts have saddled the “unruly horse” of

public policy

89

and rejected the efforts of intellectual prop-

erty producers to protect by contract that which copyright

fails to protect.

90

Contracts, then, leave databases exposed to parasitic

competitors and information Samaritans not in privity of

contract with the database maker,

91

and to the risk that a

particular court will, on public policy grounds, refuse to en-

force a private contractual agreement regarding databases.

92

3. Protection Afforded by Unfair Competition Law

Database makers who fail to obtain adequate protection

from copyright and contract can sometimes obtain protec-

tion from the misappropriation branch of unfair competition

Environment for the Publisher in the Digital World 18 (visited May 11, 1997)

<http://www.grainger.uiuc.edu/icsu/clark.htm>.

86. See, e.g., 1 E.

ALLAN FARNSWORTH, FARNSWORTH ON CONTRACTS § 1.1, at 4

n.1 (2d ed. 1990) (citing R

ESTATEMENT (SECOND) OF CONTRACTS § 1 (1981) (“A con-

tract is a promise or set of promises for the breach of which the law gives a rem-

edy . . . .”).

87. American Airlines, 115 S. Ct. at 826.

88. See Zeidenberg, 86 F.3d at 1455.

89. See 2 E.

ALLAN FARNSWORTH, FARNSWORTH ON CONTRACTS § 5.2, at 8 n.1

(2d ed. 1990) (quoting Richardson v. Mellish, 2 Bing. 229, 252, 130 Eng. Rep. 294,

303 (1824)); cf. Zeidenberg 908 F. Supp. at 657-59.

90. See, e.g., ProCD v. Zeidenberg, 908 F. Supp. 640 (W.D. Wis.), rev’d, 86

F.3d 1447 (7th Cir. 1996).

91. Perritt, supra note 4, at 292 (“contract protection is unavailable except

when privity of contract exists”).

92. See supra notes 89-90.

720 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

law.

93

The tort of misappropriation protects investors from

the malcompetitive behavior of free-riding competitors.

94

Unfortunately, the aegis of unfair competition law ebbs

and flows with the tide, a tide that follows not the moon, but

the sub rosa principles held by individual judges.

95

The re-

sulting unpredictability of the remedy renders unfair compe-

tition law less than an optimal means of protecting and

stimulating investment in databases.

96

Unfair competition law, then, while affording some pro-

tection to database makers from malcompetitive free rid-

ers,

97

does not protect database makers from information

Samaritans,

98

and does not, because of the uncertainty of a

remedy, provide a predictable basis upon which to make an

investment decision.

99

93. International News Serv. v. Associated Press, 248 U.S. 215 (1918) (giving

birth to the tort of misappropriation in the United States); Denicola, supra note

12, at 517 n.7 (collecting cases).

94. See C. Owen Paepke, An Economic Interpretation of the Misappropriation

Doctrine: Common Law Protection for Investments in Innovation, 2 HIGH TECH. L.J.

55, 68-69 (1987) (describing three common elements as: (1) investment by the

plaintiff in creating a product; (2) use of that investment or product in competi-

tion with the plaintiff; and (3) commercial damage to the plaintiff); Reichman &

Samuelson, supra note 4, at 132 (citing domestic and foreign authorities recogniz-

ing the tort); Restatement (Third) of Unfair Competition Law § 38. But see Gary

Myers, The Restatement’s Rejection of the Misappropriation Tort: A Victory for the

Public Domain, 47 S.C. L. REV. 673, 678 (1996) (arguing that the Restatement

(Third) of Unfair Competition Law rejects the tort of misappropriation as an in-

dependent cause of action).

95. See Reichman, supra note 10, at 2476.

96. Reichman & Samuelson, supra note 4, at 131; Paepke, supra note 94, at 68-

70 (noting that the tort lacks consistent formulation and application).

97. Id.; see supra note 93 (citing INS and a source collecting cases following

the INS rationale).

98. But see Paepke, supra note 94, at 68-69 (noting that some courts have

eliminated the element of competition); see generally R

ALPH S. BROWN & ROBERT C.

DENICOLA, CASES ON COPYRIGHT, UNFAIR COMPETITION, AND OTHER TOPICS

BEARING ON THE PROTECTION OF LITERARY, MUSICAL, AND ARTISTIC WORKS 580-83

(6th ed. 1995) (describing the reception and evolution of INS in the federal

courts).

99. See Samuelson, supra note 11, at 9 (noting that the drafters of the Initial

Proposal regarded EC unfair competition law as insufficient to produce adequate

incentives to invest in database production).

1997] EUROPEAN DATABASE DIRECTIVE 721

Thus, while the mosaic of these three legal regimes offers

some protection to database makers from the malcompeti-

tive or parasitic behavior of competitors, users, and informa-

tion Samaritans, existing legal regimes, both singularly and

collectively, leave database makers exposed to free riders

who can market a competing product at a fraction of the da-

tabase maker’s cost,

100

and to parasitic users and information

Samaritans who can wreak similar havoc to the database

maker’s investments.

101

As a result, many commentators

have recognized the need for additional protection to in-

crease incentives to produce databases, those valuable busi-

ness tools that place information at the fingertips of the

world,

102

without extinguishing competition, the ultimate

guardian of consumers in market economies.

103

In response to this perceived need to vindicate the com-

petitive ethos in database markets, the EC adopted the Data-

100. A database maker’s production costs can be expressed as P = dC + dP +

dM + dR, where (P) equals the sum of data collection, purchase, generation, and

validation costs (dC); plus data compiling and presentation costs (dP); plus da-

tabase marketing, distribution, billing and collection costs (dM); plus database

reproduction costs (dR). See Perritt, supra note 4, at 277 (expressing the costs of

an intellectual property originator). A database pirate’s production costs can be

expressed as p = dA + dT + dM + dR + dL; where (p) equals the sum of data ac-

quisition costs in copying from the database maker (dA); plus data transforma-

tion costs (dT); plus database marketing, distribution, billing and collection costs

(dM); plus database reproduction costs (dR); plus legal liability costs (dL). Id.

(expressing the costs of an intellectual property pirate). Assuming database

marketing (dM) and reproduction (dR) costs are equal for the maker and the pi-

rate, the pirate cannot compete in the same market with the database maker on a

price basis unless P > p, or dC + dP > dA + dT + dL. See id. However, as dis-

cussed above, head-to-head competition is not the only way a pirate can damage

the maker’s investment. See supra note 73.

101. Id.; see also supra notes 45, 73 and accompanying text (discussing sus-

ceptibility of databases to information Samaritans).

102. See supra note 38 and accompanying text; see generally Wendy J.

Gordon, On Owning Information: Intellectual Property and the Restitutionary Im-

pulse, 78 VA. L. REV. 149, 222-26 (1992) (describing a new tort of “malcompetitive

copying” when defendants knowingly copy protected intangibles in markets ex-

hibiting asymmetrical market failure that damage the creator’s market interests

without providing an equal value to the creator in the long run).

103. See supra note 28.

722 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

base Directive,

104

which prompted the United States and the

World Intellectual Property Organization to advance similar

proposals.

105

II. COMPARISON OF THE DATABASE DIRECTIVE WITH THE

PROPOSALS

This part compares the U.S. and WIPO Proposals with

the Database Directive (collectively referred to as “regimes”).

While the WIPO Proposal generally tracks the Database Di-

rective, the U.S. Proposal, H.R. 3531, differs in several sig-

nificant respects. The first apparent difference lies in the

broader definition of “database” found in H.R. 3531.

106

A. Scope of Protection

To understand the potential scope of protection afforded

by the regimes, this section first compares the regimes’ defi-

nition of “database” and the operative terms which define

and circumscribe the sui generis rights provided to database

makers.

107

This section then compares the protection af-

forded by the regimes to existing databases, and each re-

gime’s affect on existing rights and obligations.

1. Database Defined

The Database Directive and the WIPO Proposal define

“database” as a collection of independent works, data or other

materials arranged in a systematic or methodical way capa-

ble of being individually accessed by any means.

108

In con-

trast, H.R. 3531 defines “database” as “a collection, assembly

or compilation, in any form . . . of works, data or other mate-

104. See Database Directive, supra note 1, recital 1, O.J. L 77/20, at 20 (1996)

(databases are not sufficiently protected in Member States).

105. See supra notes 1, 14-20 and accompanying text.

106. See H.R. REP. NO. 3531, supra note 1, § 2.

107. See supra notes 13, 56.

108. See Database Directive, supra note 1, art. 1(2), O.J. L 77/20, at 24 (1996);

WIPO Proposal, supra note 1, art. 2(1) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

1997] EUROPEAN DATABASE DIRECTIVE 723

rial, arranged in a systematic or methodical way.”

109

H.R. 3531’s failure to require works to be independent

could sweep recordings of audiovisual, cinematographic, lit-

erary, or musical works under its protective cloak; works

that the Database Directive and the WIPO Proposal ex-

pressly disclaim.

110

Thus, initially, H.R. 3531 appears to de-

fine “database” more broadly than either the Database Direc-

tive or the WIPO Proposal.

However, H.R. 3531, like the Database Directive and the

WIPO Proposal, extends its initial cloak of protection only to

databases that are the product of a quantitative or qualitative

substantial investment of financial or other resources in the

“collection, assembly, verification, organization or presenta-

tion” of the contents of a database.

111

Therefore, although

H.R. 3531 may initially define “database” more broadly than

the Database Directive or the WIPO Proposal, the substantial

investment requirement, assuming it is interpreted similarly

under each regime,

112

reduces the potential definitional dif-

ference.

113

109. See H.R. REP. NO. 3531, supra note 1, § 2.

110. See Database Directive, supra note 1, recital 17, O.J. L 77/20, at 21

(1996); WIPO Proposal, supra note 1, Note 2.03 WIPO Doc. CRNR/DC/6 (Aug. 30,

1996).

111. See H.R. REP. NO. 3531, supra note 1, § 3(a); WIPO Proposal, supra note 1,

art. 2(4) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996); cf. Database Directive, supra

note 1, art. 7(1), O.J. L 77/20, at 25 (1996) (substantial investment in the obtain-

ing, verification or presentation of database contents).

112. With the exception of the disclaimer regarding audiovisual works

noted above, neither the Database Directive nor the Proposals define “substantial

investment.” H.R.

REP. NO. 3531, supra note 1, § 3(a); WIPO Proposal, supra note 1,

art. 2(4) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996); cf. Database Directive, supra

note 1, art. 7(1), O.J. L 77/20, at 25 (1996).

113. But see Reichman & Samuelson, supra note 4, at 98-99 (arguing that the

definition in H.R. 3531 is so broad that it might cover noncopyrightable compo-

nents of computer programs and scientific or historical theories based on fact or

data). However, H.R. 3531 flatly excludes computer programs from the scope of

its protection. See H.R. REP. NO. 3531, supra note 1, § 3(d) (“computer programs

are not subject to this Act, including without limitation any computer programs

used in the manufacture, production, operation or maintenance of a database”).

Moreover, the argument that H.R. 3531 defines “database” broadly enough to

724 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

2. Sui Generis Rights Granted

Assuming the database passes the substantial investment

threshold, the database maker

114

gains two complementary

rights: (1) the right to prohibit extraction of, and (2) the right

to prohibit reutilization of (the Database Directive), or utili-

zation of (the WIPO Proposal), or use or reuse of (H.R. 3531),

all or a substantial part of the database contents.

115

Thus, to

understand the breadth of the complementary rights, one

must examine the definitions of “extraction,” “reutilization,”

“utilization,”

116

or “use” and “reuse,” as the case may be, and

“substantial part.”

117

cover scientific or historical theories based on significant compilations of facts or

data would apply equally to the Database Directive and the WIPO Proposal.

With today’s digital technology, even three dimensional objects with appropriate

sensing devices can be reduced to a series of ones and zeros that can be compiled

and collected in a database that would find protection under each regime, pro-

vided the database meets the benchmark of protection—substantial investment.

Although segments of scientific and historical theory, along with anything else

(except computer programs) that can be reduced to a series of ones and zeros,

can be compiled in database form and protected under either regime, the re-

gimes require the second comer only to obtain those ones and zeros from other

sources, or from the database maker by fair and honest means, for example, by

purchasing a license.

114. See supra note 56.

115. The maker or rightholder is afforded two powers characteristically as-

sociated with the bundle of rights deemed property: the power to exclude and

the power to transfer. See generally Wendy J. Gordon, A Property Right in Self-

Expression: Equality and Individualism in the Natural Law of Intellectual Property,

102 Y

ALE L.J. 1533, 1546 (1993); Database Directive, supra note 1, art. 7(2), O.J. L

77/20, at 26 (1996); WIPO Proposal, supra note 1, art. 2(2), (4) WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996) (defining “extraction” and “utilization”), art. 3(1)

(granting the right to prohibit extraction or utilization); H.R. REP. NO. 3531, supra

note 1, § 4. The regimes do not create de jure property rights in data. See infra

notes 211-13.

116. Because the meaning of “utilization” as used in the WIPO Proposal is

substantially the same as the terms “use” and “reuse” as used in U.S. Proposal,

see infra note 122, for the reader’s and writer’s benefit, the term “utilization” is

not referred to in the text hereinafter.

117. The “depth” of the complementary rights (i.e., the liability of down-

stream “innocent users” who may unknowingly reuse database contents made

available by an initial infringer) is not addressed by either regime. However,

knowledge is required to impose criminal liability under H.R. 3531. See H.R.

REP. NO. 3531, supra note 1, §§ 8, 13.

1997] EUROPEAN DATABASE DIRECTIVE 725

a. The Extraction Right

The Database Directive and the Proposals define “extrac-

tion” as “the permanent or temporary transfer of all or a sub-

stantial part of the contents of a database” to another me-

dium by any means or in any form.

118

Although H.R. 3531

explicitly extends the extraction right to copies of a data-

base,

119

the extraction right provided by the Database Direc-

tive and the WIPO Proposal would also protect copies.

120

Thus, the regimes define the extraction right substantively

the same.

b. The Complementary Reutilization, or Use and

Reuse, Right

The Database Directive and the Proposals define the

complementary reutilization or use right differently in sev-

eral respects. First, the Database Directive grants a right to

prohibit only reutilization,

121

while the Proposals grant a

right to prohibit use and reuse.

122

Second, “reutilization” as

defined by the Database Directive

123

requires a positive act—

118. See Database Directive, supra note 1, art. 7(2)(a), O.J. L 77/20, at 26

(1996); WIPO Proposal, supra note 1, art. 2(2) WIPO Doc. CRNR/DC/6 (Aug. 30,

1996); H.R. REP. NO. 3531, supra note 1, § 2.

119. See H.R. REP. NO. 3531, supra note 1, § 2.

120. A copy necessarily requires transfer of the entire contents of a database,

an act prohibited by each regime. See supra note 115.

121. The Database Directive defines “reutilization” as “any form of making

available to the public all or a substantial part of the contents of a database by

the distribution of copies, by renting, by online, or other forms of transmission.”

See Database Directive, supra note 1, art. 7(2)(b), O.J. L 77/20, at 26 (1996).

122. See supra note 116. The WIPO Proposal defines “utilization” as:

[T]he making available to the public of all or a substantial part of the

contents of a database by any means, including by the distribution of

copies, by renting, or by online, or other forms of transmission, includ-

ing making the same available to the public at a place and at a time in-

dividually chosen by each member of the public.

WIPO Proposal, supra note 1, art. 2(6) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

H.R. 3531 defines “use” and “reuse” as “making available all or a substantial

part . . . of the contents of a database, or [providing] access [there]to.” See H.R.

REP. NO. 3531, supra note 1, § 2.

123. See Database Directive, supra note 1, arts. 7(1), 7(2)(b), O.J. L 77/20, at

726 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

transmission—while “use” as defined by the Proposals could

theoretically occur without a positive act.

124

Thus, based

solely on the definition of this complementary right, the

Proposals grant the database maker a broader right.

c. Breadth of the Sui Generis Rights

As noted above,

125

the breadth of the right to prevent ex-

traction and the right to prevent reutilization, or use and re-

use, is linked to the meaning of “substantial part.”

126

Here,

the WIPO Proposal diverges slightly, and H.R. 3531 diverges

significantly, from the Database Directive.

The WIPO Proposal defines “substantial part” as “any

portion of the database, including an accumulation of small

portions, that is of qualitative or quantitative significance to

the value of a database.”

127

The WIPO Proposal therefore

provides more guidance to courts charged with interpreting

it than does the Database Directive, which purposely left the

definition of “substantial part” to the European Court of Jus-

tice.

128

However, because the Database Directive specifically

prohibits repeated and systematic extraction or reutilization

of insubstantial parts of the contents of a database which

conflict with the normal exploitation of the database or

25, 26 (1996); supra note 120.

124. See WIPO Proposal, supra note 1, arts. 2(6), 3(1) WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996) (prohibiting use by making available to the public

at a time and place individually chosen by a member of the public); see also H.R.

REP. NO. 3531, supra note 1, §§ 2, 4(a)(1) (prohibiting use or reuse by providing

access).

125. See supra notes 115-17 and accompanying text (the rights to prevent ex-

traction and reutilization, or use and reuse, extends only to all or a substantial

part of the database).

126. See infra notes 127-30 (explaining the regimes’ differing definitions of

substantial part).

127. WIPO Proposal, supra note 1, art. 2(5) WIPO Doc. CRNR/DC/6 (Aug.

30, 1996).

128. See Jens-L. Gaster, The New EU Directive Concerning the Legal Protection

of Data Bases, in 3

FORDHAM INTERNATIONAL INTELLECTUAL PROPERTY LAW &

POLICY, supra note 11, at 4 (explaining that the Commission deliberately left the

task of defining “substantial part” to the European Court of Justice).

1997] EUROPEAN DATABASE DIRECTIVE 727

which unreasonably prejudice the legitimate interests of the

database maker,

129

a court could easily ascribe the same

meaning to the term, whether interpreting the term as found

in the WIPO Proposal or the Database Directive.

H.R. 3531, in contrast, implicitly defines “substantial

part” by defining “insubstantial part” as “any portion of the

contents of a database whose extraction, use or reuse does

not diminish the value of the database, conflict with a nor-

mal exploitation of the database or adversely affect the ac-

tual or potential market, for the database.”

130

Although one

could argue that the legitimate interests of the database

maker could include the potential market, the Database Di-

rective prohibits only acts which “unreasonably prejudice”

those legitimate interests,

131

while H.R. 3531 broadly prohib-

its extractions, uses, and reuses that merely “adversely af-

fect” the actual or potential market.

132

H.R. 3531 illustrates

its protective sweep by providing examples of extractions,

uses, and reuses of database contents that would constitute

infringement.

133

Some commentators have suggested that the foregoing

provisions of H.R. 3531, coupled with its severe criminal

penalties,

134

have the “potential for impeding virtually any

129. See Database Directive, supra note 1, art. 7(5), O.J. L 77/20, at 26 (1996).

130. H.R. REP. NO. 3531, supra note 1, § 2 (emphasis supplied).

131. See Database Directive, supra note 1, arts. 7(5), 8(2), O.J. L 77/20, at 26

(1996).

132. H.R. REP. NO. 3531, supra note 1, §§ 2 (defining “insubstantial part”),

4(a) (prohibited acts).

133. Infringing acts include extraction, use, or reuse of all or a substantial

part of the database contents: (1) in a product or service directly or indirectly

competing with a database from which it was extracted; (2) in a product or ser-

vice directly or indirectly competing in any market in which the database owner

has a demonstrable interest or expectation in licensing or otherwise; (3) in a

product or service for customers who might otherwise reasonably be expected to

be customers for the database; or (4) in an organization, by or for multiple per-

sons within the organization, in lieu of an additional authorized use or reuse.

H.R.

REP. NO. 3531, supra note 1, § 4(b).

134. H.R.

REP. NO. 3531, supra note 1, § 8 (providing for a fine of $250,000 or

imprisonment for five years, or both, for first-time offenders who willfully, and

728 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

judge made [‘fair use’] exceptions”

135

and, more importantly,

may cast a shadow on the availability of any pro-competitive

“Magill-like”

136

exceptions to the database maker’s “un-

bounded derivative work right.”

137

Nevertheless, despite

H.R. 3531’s potential breadth, it is narrower than the Data-

base Directive and the WIPO Proposal in two significant

ways.

138

First, H.R. 3531 expressly withholds protection for data-

bases made by governmental entities, and by governmental

for commercial advantage or who cause $10,000 damage to the database owner

in one calendar year, violate the database owners sui generis rights). The U.S.

Proposal also provides a detailed set of civil remedies for violating the database

makers’ extraction and reutilization rights. See H.R. REP. NO. 3531, supra note 1,

§ 12 (providing injunctive relief, impoundment, actual damages or statutory

damages, costs, and attorneys’ fees in certain cases, and treble damages against

persons adjudged a violator of §§ 10 or 11 within the previous three years). In

contrast, violating the database makers’ extraction and reutilization rights under

the Directive triggers “appropriate remedies.” See Database Directive, supra note

1, art. 12, O.J. L 77/20, at 27 (1996). Under the WIPO Proposal, Contracting Par-

ties must provide “expeditious remedies to prevent infringements, and remedies

that constitute a deterrent to further infringement.” WIPO Proposal, supra note 1,

art. 13 (Alternative B) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996); see also id. (Al-

ternative A incorporating the Annex).

135. Reichman & Samuelson, supra note 4, at 99.

136. Radio Telefis Eireann v. Commission, Case 69/89, [1991] E.C.R. 485, 4

C.M.L.R. 586 (Ct. First Instance); British Broadcasting Corp. v. Commission, Case

70/89, [1991] E.C.R. 535, 4 C.M.L.R. 669 (Ct. First Instance); Independent Televi-

sion Publications Ltd. v. Commission, Case 76/89, [1991] E.C.R. 575, 4 C.M.L.R.

775 (Ct. First Instance) (known collectively as “Magill”) (affirming a compulsory

licensing order against a copyright holder who used the copyright to prevent a

competitor from producing a value-added product (a television program guide);

see also

infra notes 276-92 and accompanying text (discussing other swords avail-

able in the EC).

137. See Reichman & Samuelson, supra note 4, at 101.

138. In addition, to the extent that the Commerce Clause, U.S.

CONST. art. 1,

§ 8, cl.(3), retains any vitality, it places a jurisdictional limit on Congress’s ability

to regulate databases. Cf. United States v. Lopez, 514 U.S. 549 (1995) (invalidat-

ing Gun-Free School Zones Act of 1990 as beyond Congress’s ability to regulate

under the Commerce Clause). Of course, any limit inherent in the Commerce

Clause would not affect a properly executed treaty. Cf. Missouri v. Holland, 252

U.S. 416 (1920) (unless forbidden by the United States Constitution, the federal

government can act in matters where a national interest exists that can only be

protected by acting in concert with another nation).

1997] EUROPEAN DATABASE DIRECTIVE 729

agents in their official capacities.

139

Second, H.R. 3531 insu-

lates information users from liability for the “use or reuse of

database contents lawfully extracted from a database[] prior

to the effective date of th[e] Act.”

140

3. Rights and Obligations in Existing Databases

The Proposals diverge from the Database Directive in

that the Proposals would sweep all existing databases under

their cloaks of protection,

141

while the Database Directive re-

quires protection of only those databases made after January

1, 1983.

142

The regimes converge in two respects. First, each

regime provides that the sui generis rights are without preju-

dice to other rights or obligations in a database or its con-

tents.

143

Second, neither regime extends copyright-like pro-

tection to the contents of the database. Instead, each regime

requires users or potential competitors either to collect the

information independently, or to pay the database maker for

the collected information.

144

In essence, the regimes imbue

139. H.R. REP. NO. 3531, supra note 1, § 2. However, H.R. 3531 does not ex-

clude databases merely because its contents have been obtained from a govern-

mental entity. Id. § 3(c). To accommodate this derogation, the WIPO Proposal

makes the matter of protection granted to databases made by governmental en-

tity or their agents a matter for national legislation. WIPO Proposal, supra note 1,

art. 5(2) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

140. H.R. REP. NO. 3531, supra note 1, § 15 (providing that H.R. 3531 would

become effective upon enactment). In contrast, the WIPO Proposal permits

adoption of a limited two-year window, during which time copies lawfully made

of databases before the effective date of the treaty may be distributed to the pub-

lic, see WIPO Proposal, supra note 1, art. 11(2), while the Database Directive pro-

vides only that its sui generis regime “shall be without prejudice to any acts con-

cluded and rights acquired” before the effective date. Database Directive, supra

note 1, art. 14(4), O.J. L 77/20, at 27 (1996).

141. WIPO Proposal, supra note 1, art. 11(1) WIPO Doc. CRNR/DC/6 (Aug.

30, 1996); H.R.

REP. NO. 3531, supra note 1, § 15.

142. See Database Directive, supra note 1, art. 14(3), O.J. L 77/20, at 27 (1996).

143. Id. art. 13, O.J. L 77/20, at 27 (1996); H.R.

REP. NO. 3531, supra note 1, §

9; WIPO Proposal, supra note 1, art. 12 WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

144. Database Directive, supra note 1, recital 45, O.J. L 77/20, at 23 (1996);

WIPO Proposal, supra note 1, Note 3.02 WIPO Doc. CRNR/DC/6 (Aug. 30, 1996);

H.R. REP. NO. 3531, supra note 1, § 5(b).

730 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

databases with a portable toll fence.

145

B. Beneficiaries of Protection

The Database Directive protects databases made by

146

na-

tionals and habitual residents of Member States;

147

by com-

panies and firms formed according to the law of a Member

State and having their central administration or principal

place of business within the EC;

148

and by companies and

firms having their registered office within the EC, provided

the company or firm has a genuine economic link to a Mem-

ber State.

149

The Database Directive also protects databases

manufactured in third countries by agreement of the Council

of the European Union (“Council”).

150

Similarly, the WIPO Proposal protects databases made by

nationals and habitual residents of Contracting Parties;

151

by

companies, firms, and other legal entities having their cen-

tral administration or principal place of business within a

145. See Gordon, supra note 7, at 855.

146.“Databases made by” means “database maker” as that term is defined

by the regimes. See supra notes 13, 56.

147. Database Directive, supra note 1, art. 11(1), O.J. L 77/20, at 26 (1996).

148. Id. art. 11(2), O.J. L 77/20, at 27 (1996).

149. Id.

150. Database Directive, supra note 1, art. 11(3), O.J. L 77/20, at 27 (1996).

Unless eligible under one of the first three options, or unless the United States

enacts a statutory scheme offering comparable protection,

American database

makers could not avail themselves of the sui generis regime because the final op-

tion requires third countries to provide comparable protection (material reciproc-

ity) to EC databases, Database Directive, supra note 1, recital 56, O.J. L 77/20, at

24 (1996), and because Feist prohibits United States copyright law from extending

comparable protection. See supra note 12 (discussing the Feist decision). Al-

though the material reciprocity requirement seemingly contravenes the most ba-

sic principle of the TRIPS agreement—national treatment—the Database Direc-

tive falls outside the TRIPS agreement. See Council Common Position No. 95/20,

Statement of the Council’s Reasons 19, O.J. C 288/02, at 28 (1995) [hereinafter

Common Position] (sui generis right not linked to any existing international con-

vention); Reichman, supra note 8, at 347-51.

151. See WIPO Proposal, supra note 1, arts. 6, 7(4) WIPO Doc. CRNR/DC/6

(Aug. 30,

1996).

1997] EUROPEAN DATABASE DIRECTIVE 731

Contracting Party;

152

and by companies, firms and other le-

gal entities having their registered office in the territory of a

Contracting Party, provided its operations have a genuine

and ongoing economic link to a Contracting Party.

153

H.R.

3531, in contrast, protects databases made by natural and ju-

ristic persons,

154

but, as noted above,

155

specifically excludes

governmental entities and certain governmental agents in

their official capacities.

156

C. Term of Protection

Rather than providing a single initial term of protection,

the regimes provide a two-stage initial term of protection.

157

The Database Directive and the WIPO Proposal, alternative

B,

158

provide a fifteen-year term of protection, plus an addi-

tional fifteen years if the database is made available to the

public before expiration of the initial term.

159

In contrast,

152. See id. art. 6(2), WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

153. Id. Note, however, that unlike the Database Directive, which adopts a

material reciprocity standard, see supra note 150, the WIPO Proposal adopts the

standard of national treatment by requiring Contracting Parties to grant database

makers who are nationals of other Contracting Parties the rights granted by the

Treaty and other rights granted to the nationals of the Contracting Party. See

WIPO Proposal, supra note 1, art. 7(1) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

154. See H.R. REP. NO. 3531, supra note 1, § 2 (defining “juristic person” to

exclude governmental entities).

155. See supra note 138 and accompanying text.

156. H.R.

REP. NO. 3531, supra note 1, § 2.

157. See Database Directive, supra note 1, art. 10(1)-(2), O.J. L 77/20, at 26

(1996); WIPO Proposal, supra note 1, art. 8(1)-(2) WIPO Doc. CRNR/DC/6 (Aug.

30, 1996); H.R. REP. NO. 3531, supra note 1, § 6(a).

158. See WIPO Proposal, supra note 1, art. 8(1)-(2) WIPO Doc. CRNR/DC/6

(Aug. 30, 1996) (alternative A provides a 25-year term, while alternative B pro-

vides a 15-year term).

159. The initial and second terms begin on January 1 of the year following

the date on which the database meets the requirements for protection, or is made

available to the public. See Database Directive, supra note 1, art. 10(1)-(2), O.J. L

77/20, at 26 (1996); WIPO Proposal, supra note 1, art. 8(1)-(2) WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996). The Initial Proposal limited the term of protection

to 10 years. See Initial Proposal, supra note 11, art. 9(3), O.J. C 156/03, at 9 (1992).

Apparently, the term of protection was increased to 15 years as a result of indus-

try lobbying. See Suzanne Perry, Longer Protection in Database Directive Proposed,

732 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

H.R. 3531 provides a potentially perpetual initial term of

protection,

160

plus an additional twenty-five years from the

earlier of the date the database is made available to the pub-

lic or is first placed in commercial use.

161

In addition to the initial two-stage term of protection, the

regimes provide the database maker an opportunity to ex-

tend the toll fence perpetually.

162

However, the Database

Directive and the WIPO Proposal require the same showing

to extend the fence as is required to erect the toll fence: a

substantial investment.

163

In contrast, H.R. 3531 requires

only a “change of commercial significance” to extend the

fence perpetually.

164

D. Exceptions: User Protections

The regimes protect legitimate users by providing excep-

tions to, and placing limitations on, the sui generis rights. For

example, each regime avoids conferring a de jure monopoly

REUTER EUR. COMMUNITY REP., Mar. 17, 1993, available in Westlaw, Allnewsplus

File.

160. Compare H.R. REP. NO. 3531, supra note 1, § 6(a) (term of protection is

triggered by placing in commercial use or making available to the public) with id.

§ 3(b) (databases qualifying for protection under the act remain protected regard-

less of whether they are made available to the public or placed in commercial

use).

161. See id. § 6(a) (providing that databases remain subject to the act for 25

years after January 1 of the year following the earlier of the date on which the

database is made available to the public or placed in commercial use).

162. See Database Directive, supra note 1, art. 10(3), O.J. L 77/20, at 26 (1996);

WIPO Proposal, supra note 1, art. 8(3) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996);

H.R.

REP. NO. 3531, supra note 1, § 6(b).

163. One can argue that the Database Directive and the WIPO Proposal do

not offer “renewal” options, but rather offer protection only for “new” databases

because the same investment is required to extend protection as to obtain initial

protection. See Database Directive, supra note 1, art. 10, O.J. L 77/20, at 26 (1996);

WIPO Proposal, supra note 1, arts. 8(3), 10 WIPO Doc. CRNR/DC/6 (Aug. 30,

1996).

164. See H.R. REP. NO. 3531, supra note 1, §§ 6(b), 2 (defining “change of

commercial significance” as “a change that a reasonable user of a database would

regard as affecting the quality, quantity or value” of the database contents as a

whole).

1997] EUROPEAN DATABASE DIRECTIVE 733

on facts or information,

165

and provides that all other laws

regarding databases or their contents—chief among them for

purposes of this Article, competition law and laws in respect

to access to public documents—shall remain unaffected by

the respective regime.

166

Additionally, the Database Directive and the WIPO Pro-

posal expressly permit Member States and Contracting Par-

ties, respectively, to enact limited “fair-use” exceptions.

167

H.R. 3531 does not expressly provide fair-use exceptions, but

does follow the Database Directive in permitting lawful da-

tabase users to extract, use, and reuse insubstantial amounts

of data for “any purposes whatsoever.”

168

The Database Di-

rective strengthens this user protection by deeming contrac-

tual provisions to the contrary void,

169

but qualifies the pro-

tection by prohibiting users from performing acts that either

conflict with normal exploitation of the database or unrea-

sonably prejudice the database maker’s legitimate inter-

ests.

170

In contrast, H.R. 3531 permits parties to contract

165. However, in the case of sole-source data, the regimes could result in a

de facto monopoly on facts. See infra notes 262-70 and accompanying text.

166. Database Directive, supra note 1, art. 13, O.J. L 77/20, at 27 (1996); H.R.

REP. NO. 3531, supra note 1, § 9(c); WIPO Proposal, supra note 1, art. 12 WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996).

167. See Database Directive, supra note 1, art. 9, O.J. L 77/20, at 26 (1996);

WIPO Proposal, supra note 1, art. 5(1) WIPO Doc. CRNR/DC/6 (Aug. 30, 1996).

In this respect, the WIPO Proposal would permit broader “fair-use” exceptions

than the Database Directive. Compare WIPO Proposal, art. 5(1) WIPO Doc.

CRNR/DC/6 (Aug. 30, 1996) (permitting exceptions to and limitations on the

extraction and use rights in special cases that do not conflict with the normal ex-

ploitation of the database and do not unreasonably prejudice the legitimate in-

terests of the database right-holder) with Database Directive, supra

note 1, art. 9,

O.J. L 77/20, at 26 (1996) (providing three specific enumerated exceptions which

may be authorized by Member States). The Database Directive’s exceptions are

discussed more fully infra notes 294-321 and accompanying text.

168. H.R. REP. NO. 3531, supra note 1, § 5(a); Database Directive, supra note 1,

art. 8(1), O.J. L 77/20, at 26 (1996).

169. Database Directive, supra note 1, art. 15, O.J. L 77/20, at 27 (1996).

170. Id. art. 8(1)-(2), O.J. L 77/20, at 26 (1996).

734 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

around this potential user protection.

171

H.R. 3531 also se-

verely limits “any purposes whatsoever”

172

by prohibiting

users from performing or procuring acts that singularly or

cumulatively, adversely affect the actual or potential market

for the database.

173

Considering this limitation and others

noted above,

174

“one is hard pressed to imagine”

175

a legiti-

mate purpose for which a user could extract or use even an

insubstantial amount of data from a database protected by

H.R. 3531 without the database maker’s authorization.

E. The Proposals’ Provisions Regarding Protection-Defeating

Devices and Database Management Information

The Proposals prohibit the importation, manufacture, or

distribution of devices whose primary purpose or effect is to

defeat or circumvent self-protection measures

176

taken by

database makers to prevent unauthorized extractions and

utilizations.

177

Violation of this prohibition triggers “appro-

priate and effective remedies”

178

under the WIPO Proposal,

and triggers a detailed set of civil remedies under H.R.

3531.

179

In addition, H.R. 3531 offers protection to database mak-

ers who mark their products with database management in-

171. See H.R. REP. NO. 3531, supra note 1, § 9(b) (permitting parties to freely

contract with regard to databases and their contents).

172. See supra note 168.

173. H.R.

REP. NO. 3531, supra note 1, § 4; see also supra notes 108-45 and ac-

companying notes (d

174. See supra notes 108-45 and accompanying notes (

175. See Reichman & Samuelson, supra note 4, at 100.

176. Such self-protection measures might include encryption or password-

protection systems.

177. WIPO Proposal, supra note 1, art. 10 WIPO Doc. CRNR/DC/6 (Aug. 30,

1996); H.R.

REP. NO. 3531, supra note 1, § 10. The Database Directive does not

address protection-defeating devices.

178. WIPO Proposal, supra note 1, art. 10(2) WIPO Doc. CRNR/DC/6 (Aug.

30, 1996).

179. These are the same remedies triggered for violation of the extraction

and use rights. See supra note 134.

1997] EUROPEAN DATABASE DIRECTIVE 735

formation.

180

Knowingly providing false database manage-

ment information, or knowingly altering or removing data-

base management information without authority of the da-

tabase maker, triggers both civil remedies and severe

criminal penalties.

181

III. MAKING THE PARADIGMATIC SHIFT: LINKING

INCENTIVES TO PRODUCTION COSTS

To explain the implications of the Database Directive’s

sui generis regime, this part

182

analyzes the Database Direc-

tive in light of applicable EC law, using the factual back-

ground of a recent United States case, ProCD, Inc. v. Zeiden-

berg.

183

This part also examines the balance the Database

Directive strikes in the EC between the public’s need for in-

formation access and the database maker’s need for produc-

tion incentives. Finally, this part responds to criticisms and

concerns voiced by commentators.

184

Seeking to correct the imbalance of investment between

the EC and the world, and between EC members in database

production,

185

the Database Directive seeks to harmonize ex-

isting European copyright law

186

while simultaneously erect-

ing a renewable, potentially perpetual, fifteen-year toll

180. H.R. REP. NO. 3531, supra note 1, § 11.

181. Id. §§ 11, 13 (violation of § 11 with intent to defraud can be punished by

fine of up to $500,000 or five years’ imprisonment, or both). Neither the Data-

base Directive nor the WIPO Proposal explicitly address database management

information.

182. Unless stated otherwise, this part addresses only the Database Direc-

tive, not the Proposals. So the reader will not have to refer back to Part II and

wrest information, applicable only to the Database Directive, from the Proposals,

some information previously discussed is repeated in this part.

183. 908 F. Supp. 640 (W.D. Wis. 1996), rev’d, 86 F. 3d 1447 (7th Cir. 1996).

184. See supra notes 21-23 and accompanying text.

185. See supra note 11; see also Hugh C. Hansen, International Copyright: An

Unorthodox Analysis, 29 VAND. J. TRANSNAT’L L. 579, 581 (1996) (explaining that

one purpose of the Directives is to improve European competitiveness).

186. But see supra note 5 (explaining that the Directive permits derogation

for less exacting regimes).

736 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

fence

187

that prohibits unauthorized extraction and reutiliza-

tion

188

of the whole or a substantial part

189

of electronic and

nonelectronic databases.

190

Available

191

to database makers

187. The sui generis regime creates a fence, rather than a wall, because the

regime permits some leakage (e.g., it permits insubstantial extractions and per-

mits Member States to enact ‘quasi-fair use’ exceptions). Database Directive, su-

pra note 1, arts. 7-9, O.J. L 77/20, at 25-26 (1996). Theoretically, the 15-year fence

can be extended perpetually; however, entitlement to another 15-year fence re-

quires the same showing required for initial protection: a substantial new in-

vestment. Id. art. 10(3), O.J. L 77/20, at 26 (1996). As noted above in Part II, the

initial term of protection under the Database Directive could approximate 30

years if the database maker withholds the database for almost 15 years before

initial release to the public.

188. Extraction is “the permanent or temporary transfer of all or a substan-

tial part of the contents of a database to another medium by any means or in any

form.” Database Directive, supra note 1, art. 7(2)(a), O.J. L 77/20, at 26 (1996).

Reutilization is “any form of making available to the public all or a substantial

part of the contents of a database by the distribution of copies, by renting, by on-

line or other forms of transmission.” Id. art. 7(2)(b), O.J. L 77/20, at 26 (1996).

189.“Substantial part” may be evaluated quantitatively or qualitatively.

Database Directive, supra note 1, art. 7(1), O.J. L 77/20, at 25 (1996). Though nei-

ther measure is defined by the directive, a lawful user “may not perform acts

which conflict with normal exploitation of the database or unreasonably preju-

dice the legitimate interests” of the database maker. Id. art. 8(2), O.J. L 77/20, at

26 (1996); see also Gaster, supra note 128, at (explaining that the Commission de-

liberately left the task of defining “substantial part” to the European Court of Jus-

tice); Rosler, supra note 11, at 120 (noting that many industry associations lobbied

for actual percentage levels to determine infringement).

190. Database Directive, supra note 1, art. 1(1), O.J. L 77/20, at 24 (1996). The

Initial Proposal, supra note 11, and the Amended Proposal, supra note 24, covered

only electronic databases. The Council extended the scope of the Database Di-

rective to nonelectronic databases because it obviated the need to draw a distinc-

tion between electronic and nonelectronic databases; because the Council

deemed it inappropriate for a database distributed in both electronic and

nonelectronic forms to receive different levels of protection; and because no such

distinction is drawn in other international agreements. See Common Position,

supra note 150, Statement of the Council’s Reasons 8, O.J. C 288/02, at 24 (1995).

Another need for the extension, although not noted in the Database Directive, is

that today’s high speed scanners and optical character recognition software

make electronic conversion of nonelectronic databases almost as easy as elec-

tronic conversion of electronic databases. See Samuelson

, supra note 11, at 3-4.

191. EC Member States must bring into force laws, regulations, and admin-

istrative provisions necessary to comply with the Database Directive before

January 1, 1998. Database Directive, supra note 1, art. 16(1), O.J. L 77/20, at 27

(1996). In addition, Norway, Iceland, Poland, and Turkey must implement the

Database Directive in accordance with their bilateral agreements. See Powell,

supra note 11, at 1.

1997] EUROPEAN DATABASE DIRECTIVE 737

in Member States

192

who can show a substantial invest-

ment

193

in obtaining, verifying, or presenting the underlying

data,

194

the toll fence protects the commercial interests of da-

tabase makers while attempting to accommodate the pub-

lic’s interests.

The Database Directive accommodates the public’s inter-

ests by allowing insubstantial extractions from, and reuse of,

the contents of databases made publicly available;

195

by

deeming public lending not an act of extraction or reutiliza-

tion;

196

by permitting Member States to enact “fair use” ex-

ceptions;

197

and by extinguishing the database maker’s right

to control resale after the first sale.

198

Additionally, although

the Database Directive does not mandate compulsory li-

192. See supra note 191.

193. Substantial investment in data collection can be measured qualitatively

or quantitatively. Database Directive, supra note 1, art. 7(1), O.J. L 77/20, at 25

(1996). Although the Database Directive indicates that compilation of several

recordings of musical performances on a compact disc is not a substantial

enough investment to be eligible for sui generis protection, id. recital 19, O.J. L

77/20, at 21 (1996), it fails to indicate the minimum level of investment required

for such protection.

194. Database Directive, supra note 1, art. 7(1), O.J. L 77/20, at 25 (1996), re-

cital 40, O.J. L 77/20, at 22 (1996). Obtaining/verifying data are self-explanatory

concepts that apply equally to both electronic and nonelectronic databases. Pre-

senting data is similarly self-explanatory in the context of nonelectronic data-

bases; however, the concept of “presentation” in the context of electronic data-

bases is generally inapplicable because users usually determine how such data is

presented. See Ginsburg, supra note 6; Denicola, supra note 12, at 531. Obtaining,

verifying, and presenting data will be collectively referred to hereinafter as “data

collection.”

195. Unless repeated and systematic insubstantial extractions unreasonably

prejudice the database maker’s legitimate interests. Database Directive, supra

note 1, art. 8, O.J. L 77/20, at 26 (1996); see also id. art. 15, O.J. L 77/20, at 27 (1996)

(making contracts to the contrary void).

196. Id. art. 7(2)(b), O.J. L 77/20, at 26 (1996).

197. See id. art

. 9, O.J. L 77/20, at 26 (1996) (giving Member States the option

to provide lawful users of publicly-available databases the right to privately re-

produce nonelectronic databases; the right to use for noncommercial teach-

ing/scientific uses; and the right to use for governmental security, administra-

tive, or judicial purposes).

198. Id. art. 7(2)(b), O.J. L 77/20, at 26 (1996).

738 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

censes,

199

Member States can theoretically compel them

when in the public’s interest,

200

and, realistically, European

courts seem ready to vindicate the public’s interest in com-

petition when information producers abuse their dominant

position.

201

Finally, the Database Directive contemplates the

199. Previous proposals mandated compulsory licenses on “fair and nondis-

criminatory terms” in the case of sole-source data. Sole-source data are data that

cannot be “independently created, collected or obtained from any other source.”

See Initial Proposal, supra note 11, art. 8, O.J. C 156/03, at 9 (1992); Amended

Proposal, supra note 24, art. 11, O.J. C 308/01, at 13-14 (1993).

200. Database Directive, supra note 1, art. 13, O.J. L 77/20, at 27 (1996) (“This

directive shall be without prejudice to . . . laws on restrictive practices and unfair

competition . . . .”).

201. See, e.g., Radio Telefis Eireann v. Commission, Case 69/89, [1991] E.C.R.

485, 4 C.M.L.R. 586 (Ct. First Instance); British Broadcasting Corp. v. Commis-

sion, Case 70/89, [1991] E.C.R. 535, 4 C.M.L.R. 669 (Ct. First Instance); Independ-

ent Television Publications Ltd. v. Commission, Case 76/89, [1991] E.C.R. 575, 4

C.M.L.R. 775 (Ct. First Instance). In Magill, Magill TV Guide Ltd. (“Magill”) at-

tempted to publish a comprehensive television program guide based upon

weekly listings published in separate program guides by British and Irish broad-

casters (no comprehensive listing was available to viewers in Ireland). Joined

Cases, 241/91 & 242/91, [1995] E.C.R. 743, 812, ¶ 10. Availing themselves of the

copyright protection afforded listings in Ireland and Great Britain, the broad-

casters sought and obtained an injunction against Magill, who had already

lodged a complaint with the European Commission. Id. at 812, ¶¶ 10-11. The

Commission found that the broadcasters had abused a dominant position and

ordered the broadcasters to supply their weekly listings to parties requesting

them for a reasonable rate, effectively forcing a compulsory license. Id. at 812-13,

¶¶ 12-13. The European Court of Justice upheld the compulsory licensing order,

finding that there was consumer demand for the comprehensive guide for which

there was no available substitute, and held that the broadcasters’ refusal to sup-

ply the data prevented the appearance of a new product and the broadcasters’

had therefore abused a dominant position in violation of Article 86 of the Treaty

Establishing the European Economic Community, Mar. 25, 1957, 298 U.N.T.S.

11, 1973 Gr. Brit. T.S. No. 1 (Cmd. 5179-II) [hereinafter EEC Treaty]

. Radio Telefis

Eireann, [1995] E.C.R. at 824, ¶¶ 53-54. Although Article 86 is triggered only

when abuse of a dominant position has the potential to affect trade between

Member States, David Harbord, Barriers to Entry and Exit in European Competition

Policy, 14 INT’L REV. L. & ECON. 411, 421 (1994), it is not necessary that the con-

duct in question have an actual, substantial effect on trade; rather, it is only nec-

essary the conduct is capable of having such an effect. Radio Telefis Eireann, [1995]

E.C.R. at 828, ¶ 69. Moreover, competition policy in the EC is converging. See

generally LEAH DAVISON ET AL., THE EUROPEAN COMPETITIVE MARKET (1995) (ex-

plaining that national competition policy is beginning to reflect supranational

competition policy). Furthermore, as markets within the Community become

more European in character, it will be increasingly difficult to prove that a par-

1997] EUROPEAN DATABASE DIRECTIVE 739

public’s interest in free competition by mandating a triennial

review of the sui generis regime to determine whether anti-

competitive effects require the establishment of a compul-

sory licensing scheme.

202

Yet, despite these seemingly broad protections of the

public interest, many commentators, echoing the monocular

vision of Feist Publications v. Rural Telephone Service,

203

have

expressed serious concerns about the negative effects of the

Database Directive on the “free” flow of information,

204

while others have argued that the Database Directive is anti-

competitive and will retard long-term growth in the data-

base industry.

205

To determine whether these concerns are

ticular activity will have only a purely national effect. Id. Thus, plaintiffs should

find it increasingly easier to use Articles 85 and 86 of the EEC Treaty to enjoin

anticompetitive behavior within the European Economic Community (“EEC”);

see generally Corbet, supra, note 29 (analyzing the interaction between EC compe-

tition law and intellectual property law).

202. Database Directive, supra note 1, art. 16(3), O.J. L 77/20, at 27 (1996)

(“[E]very three years . . . the Commission . . . shall examine in particular the ap-

plication of the sui generis right, . . . especially whether the application of this

right has led to abuse of a dominant position or other interference with free

competition which would justify . . . the establishment of non-voluntary licens-

ing arrangements.”).

203. See supra note 12. Feist’s myopic focus on the inability of facts to be

original (an epistemologically erroneous view, see supra note 37), and almost

singular concern over harming the public by creating a monopoly in public do-

main materials, “grossly neglect[ed] copyright’s incentive role.” See Ginsburg,

supra note 6, at 350.

204. See Rosler, supra note 11, at 146 (the Database Directive’s most danger-

ous aspect is “its potential to stifle access to information”); Reichman &

Samuelson, supra note 4, at 108-17, 123-24 (arguing that, by impeding the flow of

information, the Database Directive and Proposals will put basic science and

other public interest groups at risk); Academy Letter, supra note 23, at 1-2 (argu-

ing that the pending WIPO Proposal “would seriously undermine the ability of

researchers and educators to access and use scientific data, and would have a

deleterious long-term impact on our nation’s research capabilities”); GLOBAL

ACCESS, supra note 4, at 35 (summarizing the potential effects of an “overly pro-

tective database regime”).

205. See Rosler, supra note 11, at 108 (arguing that the Database Directive

will fail both as an international trade mechanism and as an economic policy); see

also Von Simson, supra note 23, at 766-68 (arguing that because the sui generis re-

gime will reward lower-level economic behavior (data collection), compilers will

have little incentive to invest higher-level creative activity that result in user-

friendly features). However, this view ignores the pressure to provide user-

740 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

well-founded, it is useful to apply the Database Directive’s

protections, and other offsetting ameliorative measures

available to members of the EC, to a set of facts.

A. Vindicating the Competitive Ethos in Nonproprietary

Data: Forcing Free Riders to Become Fair Followers

Aimed at increasing European market share in the data-

base market,

206

the transferable sui generis rights

207

created

by the Database Directive raise a legal barrier to misappro-

priation

208

of database makers’ investment in data collec-

tion.

209

Assuming a database maker can prove a substantial

enough investment

210

to secure protection, the sui generis re-

gime affords the maker two complementary quasi-property

friendly features inherent in the competitive market. See Digital Equip., 73 F.3d at

762 (“[P]igheaded refusal to satisfy customers’ preferences . . . leads to ruin as

rivals step in to take the business.”) In addition, it ignores the fact that higher-

level creative activity will still enjoy copyright protection under the Database Di-

rective. See generally Database Directive, supra note 1, arts. 3-6, O.J. L 77/20, at 25

(1996).

206. See supra note 11. The lagging European market can be explained in

part by fragmented nature of the European market and the economies of scale

achieved in the larger United States market. See Melnik, supra note 2, at 60; see

also Rosler, supra note 11, at 109-10 (noting that the United States retains many

advantages in the database market, including a monolingual market, and

economies of scale built upon years of industry experience and consolidation).

207. The sui generis rights may be transferred, assigned or granted under

contractual license. Database Directive, supra note 1, art. 7(3), O.J. L 77/20, at 26

(1996).

208. Database Directive, supra note 1, recital 39, O.J. L 77/20, at 22 (1996)

(providing that “this Directive seeks to safeguard the positions of makers of da-

tabases against misappropriation” of financial and human capital investment in

obtaining, verifying or presenting data); cf. International News Serv. v. Associ-

ated Press, 248 U.S. 215 (1918) (recognizing misappropriation); see generally

Wendy J. Gordon, On Owning Information: Intellectual Property and the Restitu-

tionary Impulse, 78 V

A. L. REV. 149 (1992).

209.“Data collection” is defined supra note 194.

210. See Database Directive, supra note 1, recitals 53-54, O.J. L 77/20, at 24

(1996) (burden on maker to prove substantial enough investment to secure pro-

tection). Unfortunately, like many key terms of the Database Directive, the sub-

stantial investment criterion is ill-defined. See supra note 193. Thus, the burden a

database maker must carry to secure protection under the sui generis regime will

remain uncertain until interpreted by the European courts.

1997] EUROPEAN DATABASE DIRECTIVE 741

rights:

211

the right to prevent extraction of the contents of

the database and the right to prevent reutilization of the con-

tents of the database.

212

Of capital importance in assessing

these rights is understanding, first, their linkage to the con-

tents of the database, and second, that neither extends copy-

right protection to the underlying facts and data.

213

Rather than attempting to assess these rights in a theo-

retical vacuum, the following assessment is made in light of

the factual background of ProCD v. Zeidenberg,

214

a case that

underscores the need for a sui generis regime to protect com-

pilers from the malcompetitive behavior of free riders and

information Samaritans, and provides a set of facts tailor-

made for exploring the application and outer limits of the

Database Directive’s sui generis regime.

1. The Facts of ProCD v. Zeidenberg

In creating a comprehensive, nationwide electronic

phone directory, the database maker in Zeidenberg spent ap-

211. See Gordon, supra note 115, at 1546. The sui generis regime does not cre-

ate de jure property rights in the underlying facts, although in the case of sole-

source data, which cannot be independently generated or acquired from another

source, it may create de facto property rights in the underlying facts. For this rea-

son and because of the potential exceptions and limits on the sui generis rights,

these rights lie somewhere between the exclusive property right in expression

granted by copyright and the protection afforded noncopyrightable facts by INS.

See INS, 248 U.S. at 236. With regard to nonproprietary data, the sui generis re-

gime functions more like permanent, portable trade secret protection; that is, the

second comer is presented with a two options: obtain a license from the data-

base maker to access its contents or obtain its contents by “proper means.” Cf.

Reichman, supra note 10, at 2438-40.

212.“Extraction” is defined supra at note 188; “reutilization” is defined supra

at notes 121 and 188; see Database Directive, supra note 1, art. 7(1), O.J. L 77/20,

at 25 (1996).

213. Id. recital 45, O.J. L 77/20, at 23 (1996). As a result, that which is in the

public domain, remains in the public domain. However, this postulation breaks

down in the case of sole-source data. Moreover, the incentive it creates to privat-

ize data raises concerns about the flow of information that must be addressed by

ameliorative measures. These concerns are discussed infra at text accompanying

notes 262-322.

214. See supra note 183.

742 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

proximately $10 million

215

compiling more than ninety-five

million residential and commercial listings from approxi-

mately 3,000 telephone directories.

216

After purchasing the

listings on compact disc for less than $200, the free rider elec-

tronically extracted and recompiled twenty million listings

from the maker’s database into his own database, and made

“his” listings freely

217

available over the Internet.

218

On cross

motions for summary judgment, the district court, while not-

ing the database maker’s argument that the free rider’s con-

duct was unfair and commercially destructive had “substan-

tial equitable appeal,” rejected the maker’s federal copyright

claim, as well as the maker’s state law claims of breach of

contract, misappropriation, and unfair competition.

219

Thus,

the district court’s holding left the free rider, and others like

him, free to purchase the databases of others, to ignore the

shrink-wrap license, to extract electronically the database’s

contents at a fraction of the makers’ effort, and then to sell

that content in a different wrapper for a lower price,

220

or to

215. Judge Freezes Student’s Internet Service, B. GLOBE, Sept. 24, 1995, at 41.

216. See Zeidenberg, 908 F. Supp. at 644.

217. Although the free rider charged for advertising in this case, the Data-

base Directive would prevent all free riders, whether commercially motivated or

not, from engaging in such commercially destructive (incentive-reducing) behav-

ior. See Database Directive, supra note 1, arts. 8-9, O.J. L 77/20, at 26 (1996).

218. See Zeidenberg, 908 F. Supp. at 646.

219. See id. at 643. The district court refused to treat the shrink-wrap license

as a licensing agreement. Id. at 650-51. Instead, the court held the agreement

unenforceable under the Uniform Commercial Code for lack of assent (the free

rider was a graduate student who disregarded screen warnings that use of the

product was subject to the license agreement). Id. at 655

. Although the district

court’s holding with regard to the federal copyright claim was a logical applica-

tion of Feist, the court’s rejection of the shrink-wrap license was analytically un-

sound. G. M. Hunsucker, Raising a Toll Fence to Protect the “Noncreative” La-

bors of Database Makers: The European Database Directive 15 n.43 (May 1,

1996) (unpublished manuscript, on file with Professor J. H. Reichman, Vanderbilt

University School of Law). Accordingly, the Seventh Circuit reversed on appeal.

See supra note 25; see also National Basketball Ass’n v. Motorola, Inc., 1997 WL

34001, at *34010 (2d Cir. 1997) (approving the Seventh Circuit’s analysis of the

state contract law claim).

220. See supra notes 5, 73 and accompanying text; see also Reichman, supra

note 10, at 2452-53.

1997] EUROPEAN DATABASE DIRECTIVE 743

provide that content to others free of charge.

221

2. Protecting the Database Maker’s Investment from

Free Riders and Information Samaritans

In the illustrative case, the free rider electronically ex-

tracted twenty million of the ninety-five million telephone

listings contained in the maker’s database,

222

downloaded

them to his hard drive, and then recompiled the listings into

his own database.

223

Under the sui generis regime created by

the Database Directive, the free rider’s first act violated the

database maker’s right to prevent extraction because he ex-

tracted

224

the contents of the database and because that ex-

traction was substantial.

225

Similarly, when the free rider

later made “his” database accessible via his Internet web

page, he violated the database maker’s right to prevent re-

utilization because he made a substantial part of the contents

of the maker’s database publicly available.

226

221. See supra note 5.

222. Court Decision Goes Against Pro CD, ELECTRONIC MARKETPLACE REP., Jan.

23, 1996, available in Westlaw, Allnewsplus File. To put this into perspective, if

one amortizes the data collection costs to the database maker over the all 95 mil-

lion listings, the free rider saved over two million dollars in data collection costs;

or, to put it another way, the free rider appropriated over two million dollars of

the maker’s investment in data collection.

223. See Zeidenberg, 908 F. Supp. at 645-46.

224. The free rider’s first act would have been an extraction whether he

downloaded the data to his hard drive, printer, or any other medium. See supra

note 188 (defining “extraction”). Moreover, even if the free rider had accom-

plished the transfer by manually keying the data in, and had thus expended

considerably more effort, it would still constitute an unauthorized extraction un-

der the Database Directive. See supra note 188.

225. Like many of the operative terms in the Database Directive, what pre-

cisely constitutes a “substantial part” of the contents of the database is not de-

fined. Apparently, the Commission intended to leave the definitional task to the

European courts. See Gaster, supra note 128, at . However, it would defy com-

mon sense, as well as the policy to protect database makers from the misappro-

priation of their data collection efforts, to find that the free rider, who extracted

20 million of 95 million telephone listings contained in a maker’s database did

not extract a substantial part of the contents of the database.

226. Id.

744 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Even if a court found that the free rider did not extract or

reuse a substantial part of the database contents, the data-

base maker could fall back on a second prong of protec-

tion—the right to prevent extraction or reutilization of in-

substantial parts of the database—when those repeated

insubstantial extractions or reutilizations unreasonably

prejudice the maker’s rights.

227

Here, the free rider’s data-

base permitted extraction of 1,000 listings per search, and

prior to the law suit was being accessed via the Internet ap-

proximately 20,000 times per day.

228

With the explosive growth of Internet access providers,

229

information-rich Web pages,

230

and user-friendly interfaces

that make “surfing the Net” easy even for neophytes,

231

it

seems reasonable to conclude that many users would forego

the $200 price tag of the database maker’s product in favor

of free access to the free rider’s product. Thus, by making

the contents of the database maker’s product freely available

to the public, the free rider’s actions unreasonably preju-

diced the maker’s rights and interfered with the maker’s

normal exploitation of the database. As a result, the theo-

retical EC database maker would be entitled to appropriate

remedies.

232

227. Although database makers cannot prevent lawful users from extract-

ing/reutilizing insubstantial parts of the publicly-available databases, Database

Directive, supra note 1, art. 8(1), O.J. L 77/20, at 26 (1996), lawful users must re-

frain from performing acts that unreasonably prejudice the legitimate interests of

the database maker. Id. art. 8(2), O.J. L 77/20, at 26 (1996); see also id. art. 15, O.J.

L 77/20, at 27 (1996) (making contractual provisions contrary to article 8 null and

void).

228. See Zeidenberg, 908 F. Supp. at 646.

229. Cf. id. (noting that lawsuits concerning ownership of data distributed

via the Internet is natural given the explosive growth of the Internet) (citing Jane

C. Ginsburg, Putting Cars on the “Information SuperHighway”: Authors

, Exploiters,

and Copyright in Cyberspace, 95 C

OLUM. L. REV. 1466 (1995)).

230. See Cade Metz, The 100 Top Web Sites, PC MAG., Feb. 6, 1996, at 100.

231. See Rick Ayre & Thomas Mace, Just Browsing, PC

MAG., Mar. 12, 1996,

at 100.

232. See Database Directive, supra note 1, art. 12, O.J. L 77/20, at 27 (1996)

(requiring Member States to provide appropriate remedies for infringements of

1997] EUROPEAN DATABASE DIRECTIVE 745

Thus, the Database Directive not only protects database

makers from free riders who seek to gain a competitive ad-

vantage by extracting the contents of the maker’s database

(and thereby avoid the costs of data collection),

233

but also

protects database makers from the market-destructive antics

of information Samaritans.

234

In this respect, the Directive

fills two gaps left by existing legal regimes.

235

Moreover, long-term production incentives for database

production are more sustainable under the Directive’s sui

generis regime because the existence and degree of protection

are linked to investment,

236

rather than the superficial form

of the final product hailed the linchpin of protection by

copyright law.

237

Furthermore, by jettisoning the paradoxi-

cal linchpins of arrangement and selection

238

required for da-

tabase protection under most copyright regimes,

239

the Di-

the rights provided by the Database Directive).

233. See supra note 99; see also National Basketball Ass’n, 1997 WL 34001, at

*34013-15 (rejecting the NBA’s claim that Motorola was free-riding by transmit-

ting NBA game statistics almost contemporaneously, because Motorola was not

free riding, but rather independently collecting, assembling, and transmitting the

data). In NBA v. Motorola, however, the Second Circuit hinted that its analysis

might differ if Motorola were collecting the facts (free riding) via a NBA data-

feed. Id. at *34013-14.

234. See supra note 5.

235. See supra notes 57-101 and accompanying text.

236. See Denicola, supra note 12, at 530 (sufficient production incentives can-

not be maintained without linking the existence and degree of protection to the

total production costs). Although one can doubtless provide examples of data-

bases that would be produced with far less than 15 years of protection provided

by the Database Directive, even copyright and the law of trade secrets produce

questionable results on their margins. Can one justify protecting a copyrighted

work for an additional 74 years after it has netted the creator several million dol-

lars in its first year? Eternal trade secret protection for a consumable item

proven to cause physiological damage, which has no redeeming characteristics?

237. Id.

238. The linchpins are paradoxical because although both requirements

purport to protect the public’s interest in public domain information, they also

work against the public’s interest in promoting knowledge in several ways. See

supra notes 65-71 and accompanying text; see generally Ginsburg, supra note 6, at

344-46; Reichman, supra note 10, at 2491.

239. See Berne Convention, supra note 8, art. 2(5), 828 U.N.T.S. at 221; 17

U.S.C.A. §§ 101, 103; Feist Publications, Inc. v. Rural Telephone Serv. Co., 499

746 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

rective’s sui generis regime avoids the perverse incentives

that reduce the accessibility and usability of databases

240

and

ultimately work against the public’s interest in promoting

knowledge.

241

Finally, this act of legal candor

242

in recognizing that the

compiler’s primary contribution to the promotion of knowl-

edge is through investment in data collection,

243

rather than

data arrangement or selection, leaves courts free to focus on

protecting the commercial interests of the database maker

from parasitic competitors and other misappropriators who

reduce incentives to produce commercially-valuable data-

bases.

3. Protecting the Public’s Interest in Competition

Protecting the database maker’s commercial interests,

however, does not mean that the misappropriation fence

raised by the Database Directive completely insulates the da-

tabase maker from all competition.

244

On the contrary, be-

cause the fence prohibits only extraction and/or reutilization

of the contents of databases, and does not extend copyright

protection to mere facts or data,

245

the fence merely pro-

scribes one method of data collection;

246

it does not empower

database makers to remove data from the public domain.

247

U.S. 340, 356 (1991); Database Directive, supra note 1, recitals 15-16, O.J. L 77/20,

at 21, art. 3(1), O.J. L 77/20, at 25 (1996).

240. See supra notes 65-68 and accompanying text

241. See supra notes 38, 65-71 and accompanying text

242. Reichman & Samuelson, supra note 4, at 154 (“[T]he European Commis-

sion’s frank acknowledgment that the [Database Directive’s sui generis regime]

would protect investment as such amounts to a refreshing act of legal candor.”).

243. See Denicola, supra note 12, at 530.

244. But see Reichman & Samuelson, supra note 4, at 90 (arguing that, under

the Database Directive, the database maker obtains protection exceeded only by

the protection afforded patent holders).

245. See Database Directive, supra note 1, recital 45, O.J. L 77/20, at 23 (1996).

246. See Denicola, supra note 12, at 541.

247. See supra note 213.

1997] EUROPEAN DATABASE DIRECTIVE 747

It follows necessarily that fair followers

248

remain free to

create identical databases, so long as they do not create those

databases by appropriating the contents from a protected

database.

249

Thus, under the Database Directive, the free-

rider-turned-fair-follower can go forth, collect the same data,

and compile and market an identical database or any variant

thereof,

250

without infringing the rights provided by the sui

generis regime to the database maker.

251

Alternatively, the fair follower can seek a license from the

database maker.

252

Although the Database Directive does

not mandate licensing,

253

the database maker risks price

competition from more efficient fair followers if the maker

248. See Reichman, supra note 30.

249. This statement assumes the contents of, or the database itself, do not

enjoy copyright protection. Professor Litman’s observation that this result is

functionally indistinguishable from granting copyright protection to the underly-

ing facts, Litman, supra note 37, at 1016 & n.289, holds true in the case of proprie-

tary data, but not in the case of nonproprietary data. To illustrate, assume a

copyright holder in a literary work, by definition expression not existing in the

public domain, proves that a second comer who produced an identical work had

access to the first work. Notwithstanding incantations that independent creation

is a perfect defense in copyright, see, e.g., Sheldon v. Metro-Goldwyn Pictures Corp.,

81 F.2d 49, 54 (2d Cir.), cert. denied, 298 U.S. 669 (1936), it seems incredulous to

suggest that the second comer could overcome the presumption of copying and

the inherent skepticism of judges and juries. Now assume a similar situation for

a database of nonproprietary facts protected under the sui generis regime. Here,

the second comer can more easily produce objective evidence demonstrating the

data collection costs incurred in gathering public domain materials. As a practi-

cal matter, the second comer would have a lower burden of production under

the sui generis regime.

250. This statement assumes the data are freely accessible and not otherwise

protected. Using the Zeidenberg facts as an example, if the data were protected,

the database maker would be in the same position vis-à-vis the compilers of the

3,000 directories as the free rider was to the maker. The irony of the database

maker’s suing the free rider, after cloaking himself with the constitutional pro-

tection afforded free riders by Feist (the database maker copied unprotected data

from 3,000 directories) was perhaps a, if not the, critical factor in the district

court’s decision to deny relief. See Zeidenberg, 908 F. Supp. at 658 (it is ironic that

after obtaining the benefits of copyright, the plaintiff seeks to prevent others

from obtaining those same benefits).

251. See supra notes 13, 56.

252. See supra notes 199-202 and accompanying text.

253. See infra note 273.

748 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

tries to extract too high a price for the license.

254

Similarly, because data are fungible, the database maker

who overprices his product also risks price competition from

more efficient fair followers who independently create iden-

tical databases.

255

Consequently, competitive pressures in

the nonproprietary database market should provide the da-

tabase maker with an incentive to achieve greater efficiency

in data collection: the more efficient the maker is in collect-

ing the data, the harder it will be for fair followers to com-

pete on a price basis.

256

Thus, unlike other intellectual property regimes, such as

copyright and patent, which use a winner-take-all approach

to overcome risk aversion,

257

the sui generis regime, at least

with respect to nonproprietary data, serves the same “pro-

competitive functions as does the law of trade secrets”

258

without requiring actual secrecy.

259

Therefore, with respect

to nonproprietary data,

260

the combination of market-

imposed price limits and market inducements to achieve

maximum efficiency in data collection should redound to the

benefit of the information-using public in the form of more

254. Cf. National Basketball Ass’n, 1997 WL 34001, at *34014 (explaining that

the data collector producing the cheaper or otherwise superior product will pre-

vail in the marketplace).

255. Id.

256. Other factors affect the fair follower’s ability to compete on a price ba-

sis. For example, data which is extremely difficult to collect may give the first

database maker sufficient time to gain significant market share, which in turn

may act as a barrier to entry. See supra note 100. In addition, data accuracy is

also a consideration that may affect an unknown fair follower’s ability to com-

pete because data users may be unwilling to gamble on unknown data compilers

when accuracy is critical. See Roger Elliot, Chairman’s Report (visited May 11,

1997) <http://www.grainger.uiuc.edu/icsu/confchmn.htm> (noting need for

strict peer review of scientific data submitted for publication in electronic jour-

nals); cf. Perritt, supra note 4, at .

257. See Reichman, supra note 4, at 486.

258. Zeidenberg, 86 F.3d at 1455 (citing Rockwell Graphic Sys. v. DEV Indus.,

925 F.2d 174, 180 (7th Cir. 1991)); cf. Reichman, supra note 10, at 2439-40.

259. See supra note 57.

260. See supra note 24.

1997] EUROPEAN DATABASE DIRECTIVE 749

comprehensive and less costly databases.

261

B. Criticisms of the Database Directive Regarding Sole-

Source Data: Is the Sky Falling?

June 13, 2013:

Yesterday scientists discovered a method which

stimulates the body’s immune system to a level suffi-

cient to kill the AIDS virus . . . . Unfortunately the sci-

entists have embodied these data within a database,

have set the price of each treatment at one million

dollars, and have refused to license the critical data.

—Chicken Little, The Sky Is Falling, CASSANDRA WEEKLY,

at 13.

Although perhaps a bit extreme, the hyperbole makes the

point: extending the sui generis extraction and reutilization

rights to database makers who are also data sources, or who

are the exclusive licensee of data sources, creates the risk of

data privatization,

262

with its attendant risk of monopolistic

pricing, for potentially perpetual fifteen-year renewable pe-

riods,

263

rather than just legitimately protecting the invest-

ments of database makers.

264

For that reason and several

others, it initially appears that this Article’s working hy-

pothesis, that the sui generis regime vindicates the competi-

tive ethos, breaks down in the case of sole-source data.

265

First, the implicit market limits that flow from the possi-

261. But see supra notes 204-05, infra note 362 (citing commentators who ar-

gue that regime will result in data monopolies, which will, in turn, stifle access to

information).

262. See Reichman & Samuelson, supra note 4, at 84-85 (arguing that the Da-

tabase Directive would eventually abolish public domain data).

263. See supra note 187 (noting that a “renewal” requires a substantial new

investment).

264. Even the Directive’s harshest critics agree that database makers have

legitimate interests in a new form of legal protection. See Reichman & Sam-

uelson, supra note 4, at 55; G

LOBAL ACCESS, supra note 4, at 24-25.

265. See supra notes 24, 198 (defining “sole-source data”).

750 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

bility of fair followers being more efficient data collectors are

reduced, and in limited circumstances eliminated, in the case

of sole-source data because the fence created by the sui

generis regime effectively enables the first database maker to

control dissemination of the raw data.

266

This power to con-

trol the dissemination of sole-source data enables the first

database maker not only to charge monopolistic fees for ac-

cess to the database, but also to charge monopolistic licens-

ing fees, if the database maker chooses to license at all, to fair

followers seeking to use the sole-source data in a competing

or value-added product.

267

Second, the sui generis regime creates an incentive to pri-

vatize data that in turn raises legitimate concerns over data-

base makers’ ability to control the dissemination of informa-

tion.

268

Because anyone can make a database, including

those in charge of conducting basic scientific research, this

incentive could spill over into academia and reduce the indi-

rect subsidies now provided to private research and devel-

opment.

269

Although privatizing data might result in more

short-term basic research, privatization of data as such could

stifle innovation by diminishing long-term research and de-

velopment, particularly if database makers secure exclusive

licenses and effect a cartelization of data sources.

270

The initial and amended versions of the Database Direc-

tive addressed these potential problems head-on by mandat-

ing compulsory licenses on fair and nondiscriminatory terms

266. See supra note 213

267. See Reichman & Samuelson, supra note 4, at 117-19 (sui generis regime

will impede competition in the value-added market).

268. See Rosler, supra note 11, at 141-42. Although the sui generis regimes

raise potential First Amendment issues, see, e.g., Reichman & Samuelson, supra

note 4, at 79, those issues are beyond the scope of this Article.

269. See Reichman, supra note 10, at 2498; see also infra note 356

270. Cf. Reichman, supra note 10, at 2497-98. But see infra notes 373-85 and

accompanying notes (discussing

1997] EUROPEAN DATABASE DIRECTIVE 751

in the case of sole-source data.

271

However, the Common

Position

272

removed the compulsory licensing provision in

favor of other ameliorative measures aimed at protecting the

user.

273

The next section examines these ameliorative meas-

ures in light of the criticisms regarding sole-source data

noted above,

274

to ascertain whether the Database Directive’s

sui generis regime is underprotective, overprotective, or per-

haps somewhere between the two extremes.

275

C. The Ameliorative Measures: Will They Protect the

Public’s Interest in Competition?

The Database Directive not only provides swords that

271. Initial Proposal, supra note 11, art. 8, O.J. C 156/03, at 9 (1992);

Amended Proposal, supra note 24, art. 11, O.J. C 308/01, at 13-14 (1993).

272. See supra note 150.

273. According to the Council, the compulsory licensing provision was no

longer needed in light of the narrower scope of the sui generis rights and the ex-

ceptions provided in article 9. Common Position, supra note 150, Statement of

the Council’s Reasons 15, O.J. C 288/02, at 26-27 (1995). Both justifications seem

to contradict the terms of the Database Directive. First, although the sui generis

rights, as positively stated in the Amended Proposal, extended to “part or all” of

the database, see Amended Proposal, supra note 24, art. 10(1), O.J. C 308/01, at 13

(1993), the Database Directive still prohibits repeated and systematic extraction

of insubstantial parts if that extraction conflicts with the normal exploitation of

that database or unreasonably prejudices the maker’s rights. See Database Direc-

tive, supra note 1, art. 7(3), O.J. L 77/20, at 26 (1996); see also id. art. 8(2), O.J. L

77/20, at 26 (1996) (prohibiting users from performing any act; e.g., tampering

with the database, that conflicts with the normal exploitation of that database or

unreasonably prejudices the maker’s rights). Second, the “fair use” exceptions in

article 9 are optional within the Member States. See id. art. 9, O.J. L 77/20, at 26

(1996). Nonetheless, the Council’s citation to them as a reason for removing the

compulsory license indicates its intent that the exceptions be passed. See Com-

mon Position, supra note 150, Statement of the Council’s Reasons 15, O.J. C

288/02, at 26-27 (1995). Accordingly, even if a particular member state did not

enact the exceptions, they would probably be significant factors in any appeal to

the Commission regarding abuse of the sui generis rights. See infra notes 277-92

and accompanying text (discussing

274. See supra notes 262-70 and accompanying text.

275. For example, a solution that approximates pareto optimality. “Pareto

optimality” refers to the point at which no change can be implemented that will

make someone better off without making someone else worse off. See VISCUSI,

supra note 38, at 74-75.

752 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

enable fair followers to protect the public’s interest in com-

petition, but also provides shields to users who are not seek-

ing to compete, but are performing functions deemed suffi-

ciently important to provide exceptions to the sui generis

rights afforded database makers.

276

While examining the

swords provided by the Database Directive, to fully under-

stand the potential swath fair followers can cut with these

swords, one must simultaneously examine the arenas in

which they can be drawn and the decision-makers who will

determine whether these swords are required to protect the

public’s interest in competition in a given case.

1. The Swords: The European Economic

Community Treaty and Competition Law

Once fully implemented, the Database Directive’s sui

generis regime will, like other laws passed pursuant to the

European Economic Community (“EEC”) Treaty,

277

take

precedence over national laws.

278

However, the sui generis

rights are expressly limited by both national and Commu-

nity rules of competition.

279

As a result, a fair follower ag-

grieved by the anticompetitive behavior of a database maker

can file a complaint not only with the Commission of the

European Communities (“Commission”)

280

if the anticom-

petitive behavior violates one of the EEC competition

276. See infra notes 294-318 and accompanying text.

277. Treaty Establishing the European Economic Community, Mar. 25,

1957, 298 U.N.T.S. 11, 1973 Gr. Brit. T.S. No. 1 (Cmd. 5179-II).

278. See Corbet, supra note 29, at 330.

279. See Database Directive, supra note 1, recital 47, O.J. L 77/20, at 23 (1996).

The EC rules of competition are reported in Corbet, supra note 29, at note 84.

280. The Commission investigates and adjudicates complaints under Arti-

cles 85 and 86 that address the distortion of competition and the abuse of a

dominant position. Corbet, supra note 29, at 332. The Commission’s decision is

appealable to the Court of First Instance of the European Communities, and

questions of law are thereafter appealable to the Court of Justice of the European

Communities. Id.; see also Radio Telefis Eireann, [1995] E.C.R. at 828, ¶¶ 67-68 (ap-

peal to Court of Justice only on questions of law, not facts).

1997] EUROPEAN DATABASE DIRECTIVE 753

rules,

281

but can also file a complaint in a national forum

based on violation of the EEC Treaty

282

or any national com-

petition rule.

283

Perhaps the most powerful sword a fair follower can

wield is the one provided by Article 86 of the EEC Treaty

(“Article 86”). Article 86 prohibits the abuse of a dominant

position within the common market, or in a substantial part

of the common market, that affects trade between Member

States.

284

Value-added competitors have effectively wielded

the sword afforded by Article 86 to force compulsory li-

censes even when copyright law buttressed the dominant

position.

285

For example, in Radio Telefis Eireann v. Commission

(“Magill”),

286

the European Court of Justice upheld a com-

pulsory licensing order forcing television broadcasters to

supply their weekly program listings to a competitor who

sought to produce a value-added product (a television pro-

gram guide) not otherwise available on the market.

287

If the

European Court of Justice is willing to vindicate the public’s

interest in competition for television program guides so the

281. See Corbet, supra note 29, at 347. The EEC competition rules include

articles 85-94 of the EEC Treaty, and other treaties. Id. at 348 n.84. Articles 85

and 86 of the EEC Treaty are by far the most important, id., and are the only ones

of concern in this Article.

282. Although national forums may apply EEC law, they must refer ques-

tions of EEC law to the European Court of Justice for an interlocutory and bind-

ing (in the case and as precedent) ruling. See Corbet, supra note 29, at 330-32.

283. In this regard, the Commission has noted the tendency of Member

States to redraft their national rules consistent with EEC law. See D

AVISON ET AL.,

supra note 201, at 46-47.

284. See EEC Treaty, supra note 201, art. 86, 298 U.N.T.S. 11, 1973 Gr. Brit.

T.S. No. 1.

285. See Radio Telefis Eireann, [1995] E.C.R. at 823, ¶ 50 (noting that failure to

supply information prevented new product from appearing on the market). As

the Magill court explained, “it is . . . clear . . . that the exercise of an exclusive

right by the proprietor may, in exceptional circumstances, involve abusive con-

duct.” Id. (citing Volvo AB v. Erik Veng (UK) Ltd., Case 238/87, [1988] E.C.R.

6211, [1988] 4 C.M.L.R. 122)).

286. Joined Cases, 241/91 & 242/91, [1995] E.C.R. 743; see supra note 136.

287. See Radio Telefis Eireann, [1995] E.C.R. at 837, ¶ 104, [1995].

754 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

public may “arrange [their] leisure activi-

ties . . . accordingly,”

288

how much more willing will the

Court be to vindicate that interest in areas vital to scientific

and technical progress?

In addition, to attack database makers who secure exclu-

sive licenses, or otherwise engage in collusive activity to

thwart competition, fair followers can wield the sword pro-

vided by Article 85 of the EEC Treaty (“Article 85”).

289

Taken together, Articles 85 and 86, and the “Excalibur-like”

pro-competitive sword found in Magill,

290

along with the in-

creasing willingness of European firms to abandon anticom-

petitive practices when challenged by the Commission,

291

may be sufficient in and of themselves to protect the public’s

interest in competition by forcing would-be data monopoliz-

ers to supply their data on reasonable terms.

292

Yet the Da-

tabase Directive goes further by mandating a triennial re-

view of the sui generis rights to determine whether the

public’s interest in free competition requires additional pro-

tective measures, including additional swords such as com-

pulsory licenses, or additional shields such as the ones de-

scribed below.

293

288. Id. at 795, ¶ 173.

289. Article 85 prohibits “all agreements between undertakings . . . and con-

certed practices” which may prevent, restrict, or distort competition within the

common market. See EEC Treaty, supra note 201, art. 85, 298 U.N.T.S. 11, 1973

Gr. Brit. T.S. No. 1.

290. The swords available to fair followers in the EC are preserved by

TRIPS. See Agreement on Trade-Related Aspects of Intellectual Property Rights,

33 I.L.M. 1197, in General Agreement on Tariffs and Trade: Multilateral Trade

Negotiations Final Act Embodying the Results of the Uruguay Round of Trade

Negotiations, Apr. 15, 1994, 33 I.L

.M. 1125, Annex 1C, art. 40 (2) [hereinafter

TRIPs]; see also Fox, supra note 29, at 485 (“TRIPS reserves to nations the right to

control anticompetitive practices.”).

291. See DAVISON ET AL., supra note 201, at 46.

292. The Commission, whose proposals contained a compulsory licensing

requirement, see the Initial and Amended proposals, supra notes 11, 24 will be

available to the fair follower as an initial adjudicative body capable of compel-

ling the license. See Corbet, supra note 29, at 322-23.

293. See Database Directive, supra note 1, art. 16(3), O.J. L 77/20, at 27 (1996).

The drafters of the Database Directive intended that the same pro-competitive

1997] EUROPEAN DATABASE DIRECTIVE 755

2. Mandatory and Optional Shields to Protect the

User

Unlike the swords afforded fair followers in the interests

of competition, the protections discussed in this section pur-

port to act as shields to protect the legitimate rights of users

and others.

294

The Database Directive mandates two shields

and permits Member States to enact three “fair use” shields.

Although the Council

295

based its decision to eliminate the

compulsory licensing provision in part on the “fair use”

shields,

296

their adoption remains optional to Member

States.

297

a. The Mandatory Shields

The first mandatory shield deems public lending not an

act of extraction or utilization. Logically, this shield would

principles be available to fair followers seeking to compete with sui generis

rightholders. See Database Directive, supra note 1, recital 47, O.J. L 77/20, at 23

(1996). As explained in the Database Directive:

[I]n the interests of competition between suppliers of information prod-

ucts and services, protection by the sui generis right must not be af-

forded in such a way as to facilitate abuses of a dominant position, in

particular as regards the creation and ditribution [sic] of new products

and services which have an intellectual, documentary, technical, eco-

nomic or commercial added value . . . .

Id. (emphasis added).

294. Database Directive, supra note 1, recitals 49-51, O.J. L 77/20, at 23

(1996).

295. See supra note 150 and accompanying text.

296. See Common Position, supra note 150, Statement of the Council’s Rea-

sons 15, O.J. C 288/02, at 26-27 (1995) and Database Directive, supra note 1, art. 9,

O.J. L 77/20, at 26 (1996).

297. See Common Position, supra note 150, Statement of the Council’s Rea-

sons 15, O.J. C 288/02, at 26-27 (1995) and Database Directive, supra note 1, art. 9,

O.J. L 77/20, at 26 (1996). The Parliament’s failure to require adoption of these

optional shields conflicts with the professed harmonization goals of the EC in

promulgating the Database Directive. See Common Position, supra note 150,

Statement of the Council’s Reasons 4, O.J. C 288/02, at 23 (1995). Failure to

mandate adoption of these shields by Member States could work especial hard-

ships on academicians and scientists whose research knows no artificial territo-

rial boundaries. See SCIENCE & ENGINEERING INDICATORS, supra note 23, at 7-4 (ex-

plaining that scientific and academic research is an international undertaking

that knows now artificial geographic or political boundaries).

756 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

belong to libraries, but it appeared in the Common Position

without comment or explanation.

298

The second mandatory shield permits users to make in-

substantial extractions and reutilizations for “any purposes

whatsoever.”

299

“Any purposes whatsoever” is limited by

the correlative obligations imposed on users to refrain from

committing acts that “conflict with a normal exploitation of

the database or unreasonably prejudice the legitimate inter-

ests of the maker of the database.”

300

Given the unabashed

commercial focus of the Database Directive on the database

maker’s investment interests,

301

users hiding behind this

shield should tread carefully when their use might prejudice

those commercial interests.

On the other hand, the commercial focus of the Database

Directive

302

might aid those using database contents for non-

commercial purposes. For example, a scientist recompiling

insubstantial portions of old data for noncommercial pur-

poses, from a database whose commercial value derives

primarily from the timeliness of its data,

303

could argue that

298. See Database Directive, supra note 1, art. 7(2), O.J. L 77/20, at 26 (1996).

However, logically library users would still be subject to limits of the Database

Directive; else, this provision would allow competitors to circumvent the protec-

tions afforded by the sui generis regime by accessing databases through libraries.

299. See id. art. 8(1), O.J. L 77/20, at 26 (1996); see also id. art. 15, O.J. L 77/20,

at 27 (1996) (making contrary contractual provisions null and void). Apparently,

defining what constitutes an “insubstantial part” is a definitional task entrusted

to the European Court of Justice. See Gaster, supra note 128, at

.

300. Database Directive, supra note 1, art. 8(2), O.J. L 77/20, at 26 (1996); cf.

Berne Convention, supra note 8, art. 9(2), 828 U.N.T.S. at 238-39.

301. Database Directive, supra note 1, recitals 7-12, 38-42, 47, O.J. L 77/20, at

20, 22-23, art. 7, O.J. L 77/20, at 25 (1996); see also Common Position, supra note

150, Statement of the Council’s Reasons 14, O.J.

C 288/02, at 26 (1995).

302. Database Directive, supra note 1, recitals 7-12, 38-42, 47, O.J. L 77/20, at

20, 22-23, art. 7, O.J. L 77/20, at 25 (1996); see also Common Position, supra note

150, Statement of the Council’s Reasons 14, O.J. C 288/02, at 26 (1995).

303. Note, however, that the value/price of data varies based on factors

other than just timeliness. For example, value-added features such as analyses,

verification, or accompanying analytical tools increase the price of data. In addi-

tion, data, like any other commodity, are subject to basic economic principles

such as the law of supply and demand. See generally ADAM SMITH, THE WEALTH

1997] EUROPEAN DATABASE DIRECTIVE 757

such activity does not prejudice the database maker’s legiti-

mate, commercial interests.

Moreover, the commercial focus of the Database Direc-

tive may convince the European courts to examine the na-

ture of the commercial interests affected when determining

whether an extraction or reutilization is “substantial.”

304

For

example, a user extracting one week of old data can better

argue that such an extraction was legitimate, or insubstan-

tial, than can a user who extracted one week of current,

timely data.

305

Thus, the commercial focus of the Directive, coupled

with the Commission’s decision to leave the task of defining

“substantial part” to the European Court of Justice,

306

per-

mits the European courts to view users through a commer-

cial lens to determine whether their extraction or reutiliza-

tion prejudices the database maker’s legitimate, commercial

interests.

b. The Optional “Fair Use” Shields

The remaining “fair use” shields are optional and cannot

be exercised in such a way as to prejudice the database

maker’s rights.

307

Some commentators argue that even if

Member States adopt these optional shields, the Database

Directive still fails to protect those advancing fundamental

science.

308

OF NATIONS 49 (1776). For example, raw, real-time financial data (equity quotes)

can be obtained from several sources (such as CNNfn) free of charge (if one has

Internet access). However, because over time these same data become scarce

(only those who capture and store the real-time data will possess the data), the

supply curve shifts and, consequently, assuming the demand curve does not

shift proportionately in a corresponding direction, the price of these data in-

creases. See generally M

ILTON H. SPENCER, CONTEMPORARY MACROECONOMICS 43-

56 (7th ed. 1990).

304. See supra notes 193, 301.

305. But see note 303.

306. See supra note 128.

307. Database Directive, supra note 1, recital 50, O.J. L 77/20, at 23 (1996).

308. See G

LOBAL ACCESS, supra note 4, at 13-14 (arguing that these shields are

758 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Assuming the Member States adopt these shields, they

collectively allow extraction of substantial parts of publicly-

available databases for private purposes in the case of

nonelectronic databases, and for illustration in teaching and

for scientific research, provided the user indicates the source,

to the extent justified by the noncommercial purpose; and

extraction and reutilization of substantial parts for public se-

curity, administrative, or judicial purposes.

309

The first optional shield, which permits extraction of sub-

stantial parts of publicly-available databases for private pur-

poses,

310

illogically distinguishes between nonelectronic and

electronic databases. By using high-speed scanners and op-

tical character recognition software, nonelectronic databases

can be electronically converted almost as easily as electronic

databases.

311

Moreover, the reasons given by the Council for

extending the scope of the Database Directive to include

nonelectronic databases apply equally here, but apparently

received no consideration.

312

The second optional shield, which permits extraction of

substantial parts of publicly-available databases for illustra-

tion for teaching or for scientific research,

313

could go far in

ameliorating the concerns of those decrying the Database Di-

rective’s sui generis regime because of its potential effect on

research.

314

Provided the user indicates the source, and can

justify the use as noncommercial, only the database maker’s

legitimate interests limit the user’s ability to use the contents

of the database for the “purposes of illustration for teaching

of little use to those conducting scientific research).

309. Database Directive, supra note 1, art. 9, O.J. L 77/20, at 26 (1996). The

third and final optional shield, permitting not only extraction, but also reutiliza-

tion of substantial parts of the database, in the interests of public security, and

for the purposes of administrative or judicial procedures, is not discussed herein.

310. Id. art. 9(a), O.J. L 77/20, at 26 (1996).

311. See supra note 190.

312. See Common Position, supra note 150, Statement of the Council’s Rea-

sons 8, O.J. C 288/02, at 24 (1995).

313. Database Directive, supra note 1, art. 9(b), O.J

. L 77/20, at 26 (1996).

314. See supra note 23.

1997] EUROPEAN DATABASE DIRECTIVE 759

or scientific research.”

315

Some argue that this second optional shield is merely

“fool’s gold,” which reinforces the overly protective thrust of

the Database Directive.

316

Under this view, the scientist can

illustrate his or her conclusions, but cannot browse the data-

base or use the collected data without succumbing to the

monopoly power of the sole-source database maker. Al-

though one could read the Database Directive in this man-

ner, a careful, contextual reading supports the previous

view:

317

noncommercial uses of the database contents for the

purposes of illustration for teaching or for scientific research

is limited only by the legitimate interests of the database

maker.

318

Even if the European courts follow the parsimonious

view of the second optional shield, the fact remains that

anyone, including individual academicians or collective sci-

entific and technical institutions, can become database mak-

ers.

319

This fact should not be dismissed lightly, either by le-

gal or economic theorists, or by commercial data compilers,

because, by wielding the rights afforded by the Database Di-

rective’s sui generis regime, the information-producing aca-

315. Database Directive, supra note 1, art. 9, O.J. L 77/20, at 26 (1996); see su-

pra note 305.

316. See Reichman & Samuelson, supra note 4, at 89-90.

317. Id. The latter view ignores the difference between the shield provided

in the copyright domain and the shield provided in the sui generis domain. Com-

pare Database Directive, supra note 1, art. 6(2)(b), O.J. L 77/20, at 25 (1996) (“for

the sole purpose of illustration for teaching or scientific research”) (copyright do-

main) with Database Directive, supra note 1, art. 9(b), O.J. L 77/20, at 26 (1996)

(for the purposes of illustration for teaching or scientific research) (sui generis do-

main) (emphasis added)

; cf. Gaster, supra note 128, at (noting that the excep-

tions “usually only relate to the extraction right . . . because re-utilization is pre-

dominantly of a commercial nature”).

318. Logically, the database maker’s legitimate interests will be broader vis-

à-vis the academic or scientist if the database is designed for, and marketed to,

the academic or scientific communities. However, the scientist or academic can

still resort to the mandatory shields discussed above.

319. See supra note 13.

760 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

demic and scientific communities

320

could potentially be-

come gatekeepers to the technical and scientific data mar-

ket.

321

The significance of the scientific and academic communi-

ties fending for themselves in a commercialized information

market is considered more fully in the context of the oppor-

tunities and threats to basic science discussed in Part IV of

this Article. For now, it is sufficient to note that even if the

second optional shield proves to be “fool’s gold,” the swords

and shields available in the EC should ensure that the toll

fence raised by the Database Directive remains a pro-

competitive fence, not an insuperable monopolistic wall.

322

D. Effects within the European Community

Although the shadow of protection cast by the Database

Directive’s sui generis regime will remain uncertain until in-

terpreted by the authoritative administrative and judicial

bodies, to the extent that legal protections create incentives

320. See, e.g., SCIENCE & ENGINEERING INDICATORS, supra note 23, at 5-30 (not-

ing that, in 1993 alone, 414,000 articles were published in the 4,681 peer-reviewed

natural science and engineering journals recognized in the Institute of Scientific

Information’s Scientific Citation Index).

321. Indeed, the scientific community has begun to assume this role in some

scientific disciplines. See, e.g., Paul Ginsparg, Winners and Losers in the Global Re-

search Village 3 (visited May 11, 1997) <http://www.lanl.gov/blurb/

pg96unesco.html> (noting that automated online archives of physics research,

contributed by the scientific community, has virtually supplanted conventional

journals); see also J. C. Sens, Electronic Publishing via Scientific Societies (visited

May 11, 1997) <http://www.grainger.uiuc.edu/icsu/sens.htm> (outlining a

proposal for a non-profit organization to provide unreviewed scientific articles

free of charge, and to provide reviewed articles at a cost sufficient to finance the

non-profit organization). If this trend continues, the “freedom of information”

position taken by some in the academic and scientific communities will be put to

the test, and we shall see precisely how “free” those communities believe infor-

mation should be.

322. But cf. Reichman & Samuelson, supra note 4, at 90 (arguing that the pro-

tection afforded by the Database Directive to data is exceeded only by the protec-

tion granted by the classic patent paradigm); see also GLOBAL ACCESS, supra note

4, at 35 (arguing that under “an exclusive rights model . . . a database owner[]

[has an] absolute monopoly . . . .”).

1997] EUROPEAN DATABASE DIRECTIVE 761

to produce,

323

the sui generis regime should increase database

production in the EC.

324

At the same time, the United States

Supreme Court has left diligent compilers in the United

States vulnerable to the misappropriative efforts of parasitic

competitors and information Samaritans.

325

Unless the

United States adopts a regime providing equivalent protec-

tion, the synergistic effect of these two factors, coupled with

the increasing convergence of the European market, may in-

deed help the EC achieve their goal of gaining a larger share

of the worldwide database market.

326

With respect to nonproprietary data, the Database Direc-

tive’s sui generis regime should increase information avail-

ability and accessibility within the EC by increasing produc-

tion incentives.

327

In addition, because protection is linked

to production costs, rather than the superficial form of the

final product, the Database Directive’s sui generis regime

should operate without the price-distorting effects of an en-

hanced copyright regime.

328

While increased information availability and accessibility

under the Database Directive’s sui generis regime will come

at a certain price, absent adequate production incentives in-

323. Perritt, supra note 4, at 261 (citing Locke and Blackstone); Karjala II, su-

pra note 38, at 2594 (explaining that an absence of legal protection for intellectual

property acts as a disincentive to invest); Lunney, supra note 64, at 486 n.5 (prop-

erty rights are used to provide incentives to invest in intellectual property pro-

duction); Ginsburg, supra note 6, at 349.

324. See Samuelson, supra note 11, at 6 (pointing out that the two countries

with the biggest market share have historically protected “noncreative” labors of

database makers through “sweat of the brow” and “industrious collection” ra-

tionales). However, the Database Directive’s sui generis regime could reduce in-

centives in Great Britain where ‘sweat-of-the-brow’ databases already enjoy pro-

tection under copyright law. See supra note 8.

325. See Feist Publications, Inc. v. Rural Telephone Serv. Co

., 499 U.S. 340

(1991); supra note 10 (discussing information Samaritans).

326. See Database Directive, supra note 1, recitals 11-12, O.J. L 77/20, at 20

(1996).

327. See supra notes 236, 322-23.

328. See supra notes 67-68 and accompanying text; see also Denicola, supra

note 12, at 530.

762 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

formation seekers still pay a price: instead of paying a fee to

access an available database, information seekers pay in the

form of lost-opportunity costs—costs incurred in the form of

time and effort expended in finding the information,

329

as-

suming the information can be found at all. Thus, in the case

of nonproprietary data, the Database Directive will give in-

formation seekers an otherwise unavailable choice: pay the

gatherer for her efforts and marshal their energies toward

uses they deem more valuable, or gather their own data.

330

In contrast, in the case of some proprietary or sole-source

data, information seekers are not given a choice—they can-

not practicably go forth and gather their own data.

331

How-

ever, the swords and shields available in the EC may inject

enough legal uncertainty in the minds of would-be data mo-

nopolizers to encourage them to engage in reasonable be-

havior vis-à-vis competitors and others seeking licenses.

332

In

light of these mechanisms, and the Database Directive’s

mandated triennial assessment of the sui generis rights con-

cerning abusive and anticompetitive behavior,

333

concerns

regarding the Database Directive’s anticompetitive effects

and effect on information flow within the EC,

334

while not

eliminated, are probably overstated.

335

IV. THE DATABASE DIRECTIVE: STEPPING STONE TO AN

329. See generally A. MITCHELL POLINSKY, AN INTRODUCTION TO LAW AND

ECONOMICS (2d ed. 1989).

330. Because of the explosive growth of the Internet the latter choice is be-

coming increasingly available. See supra notes 230-31.

331. See supra note 199 (defining “sole-source data”).

332. Cf. Reichman & Samuelson, supra note 4, at 149 (arguing that the legal

uncertainty arising from fair use requirements gives producers and users

“maximum incentives to negotiate . . . licenses” on reasonable terms).

333. See Database Directive, supra note 1, art. 16, O.J. L 77/20, at 27 (1996);

see also id. recital 47, O.J. L 77/20, at 23 (1996).

334. See supra notes 22-23, 205 and accompanying text.

335. It seems more plausible to assume that the EC will respond to problems

associated with the Database Directive, than to assume that it will continue

shooting itself in the foot after a problem is detected.

1997] EUROPEAN DATABASE DIRECTIVE 763

INTERNATIONAL MODEL

This part first proposes that any international sui generis

regime modeled on the Database Directive be modified by

incorporating a compulsory licensing mechanism to enable

competition in the case of sole-source data. This part then

examines the purported hazards lurking, and the opportuni-

ties awaiting, the scientific and academic communities in an

environment that privately allocates information based upon

price.

A. Enabling Competition in a Commercialized Information

Market

Collectively, the Database Directive, along with suprana-

tional and national European competition laws (“European

competition law”), provide ameliorative mechanisms suffi-

cient to offset, or permit downstream adjustments to com-

pensate for, the potential problems associated with a regime

which may permit de facto, although not de jure, ownership

rights in sole-source data.

336

However, similar ameliorative

mechanisms, like the pro-competitive “Excalibur-like” sword

found in Magill and other pro-competition policies embed-

ded in European competition law,

337

are not available at an

international level.

338

Consequently, an international sui

generis regime may require a mechanism to ensure that new

and value-added information products are not kept off the

market by those who refuse to license sole-source data.

339

336. See supra notes 211-13 and accompanying text.

337. See supra notes 277-92 and accompanying text.

338. See, e.g., Fox, supra note 29, at 487-90 (noting the lack of any coherent

international competition law and discussing five scenarios where the two major

models, United States antitrust law and EC competition law, would likely differ

in application).

339. See, e.g., Reichman & Samuelson, supra note 4, at 77.

764 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

1. Mandating Licenses in the Case of Sole-Source

Data

One such mechanism, found in the sui generis regime ini-

tially proposed by the EC, is a compulsory licensing

scheme.

340

A compulsory licensing mechanism could be

loosely modeled on the compulsory licensing mechanism

copyright provides for nondramatic musical works.

341

Prop-

erly designed, such a mechanism could enhance competition

in a fully commercialized information market by ensuring

that value-added products are not kept off the market by da-

tabase makers who refuse to license sole-source data, and by

ensuring that the specter of competition hovers as a potential

restraint on those database makers inclined to seek monopo-

listic rents for sole-source data.

2. Compulsory Licenses Should Serve the Interests

of the Database Maker and the Public

A compulsory licensing mechanism could be designed

with a dual rate structure, one based on a statutory mini-

mum rate, the other based upon the database maker’s actual

costs. Such a rate structure could protect the public’s inter-

est in competition in the sole-source data market without

unduly prejudicing the commercial interests of the database

maker.

As a starting point, the statutory rate could be deter-

mined by multiplying the number of bytes

342

reused or ex-

340. See Initial Proposal, supra note 11, art. 8, O.J. C 156/03, at 9 (1992);

Amended Proposal, supra note 24, art. 11, O.J. C 308/01, at 13-14 (1993) (mandat-

ing compulsory licenses on “fair and nondiscriminatory terms” in the case of

sole-source data); cf. Reichman, supra note 10, at 2539-44 (proposing off-the-rack

liability regime with built-in compulsory license); see also Samuelson et al., supra

note 35, at 2412-14.

341. See, e.g., 17 U.S.C.A. § 115 (West Supp. 1996).

342.“Byte” is a term used to quantify data streams. See, e.g., THE AMERICAN

HERITAGE COLLEGIATE DICTIONARY 194 (3d ed. 1993) (defining a “byte” as equiva-

lent to one character or eight bits).

1997] EUROPEAN DATABASE DIRECTIVE 765

tracted times a small charge, such as a fraction of a cent, or

other appropriate monetary unit.

343

Alternatively, database

makers should remain free to reject the statutory rate in fa-

vor of a second rate structure based upon actual costs.

Under this second rate structure, the database maker

should bear the burden of proving actual costs associated

with particular data, but could thereafter charge accord-

ingly.

344

This rate structure would legitimately enable data-

base makers with high data-collection costs, or narrow mar-

kets over which to spread their data-collection costs, to

recover a greater proportion of their costs from would-be

competitors.

343. Cf. 17 U.S.C.A. § 115 (covering mechanical copyright royalties). A fur-

ther refinement of the compulsory licensing scheme might include a prohibition

on substantial electronic extraction and conversion by competitors for a short

blocking period designed to provide the maker with sufficient lead-time to re-

cover her costs in the market. Cf. Samuelson et al., supra note 35, at 2412-14

(blocking period followed by compulsory license); see also Reichman, supra note

10, at 2544-51. In addition, a compulsory licensing mechanism might encourage

formation of private collection societies which would permit interested parties to

bargain around the statutory rate. Cf. Reichman & Samuelson, supra note 4, at

140 (noting that the Harry Fox agency acts as a “de facto collection society

[for] . . . some 200,000 voluntary licenses . . .”); cf. Kennedy & Dweck, supra note

53, at C17 (reporting that the Author’s Guild and the American Society of Jour-

nalists formed the Author’s Registry this year to represent freelance writers in an

effort to impose usage-based royalties on publishers for electronic uses of their

works). Professors Reichman and Samuelson suggest that a statutory rate be set

by industry in light of two criteria: (1) the cost-recovery principle familiar to sci-

entific agencies, which focuses on the reasonableness of profit margins, and (2)

the value the licensed data adds to the fair follower’s product. See Reichman &

Samuelson, supra note 4, at 140. They also suggest that a statutory rate would

result in bargaining by users and producers, and would permit “a more nuanced

schedule of royalties . . . than a statute could . . . institute.” Id. at 141.

344. Although the database maker might incur additional costs in proving

actual costs, proving them should be less costly than triggering competition law

because the maker need only produce reliable evidence of her data collection

costs. Moreover, since each sui generis regime saddles the database maker with

the burden of proving substantial investment, the prudent database maker will

have adequate documentation of investment costs. See Database Directive, supra

note 1, art. 7(1), O.J. L 77/20, at 25 (1996); H.R.

REP. NO. 3531, supra note 1, §§ 2,

3(a), 6(a); WIPO Proposal, supra note 1, Note 2.07 WIPO Doc. CRNR/DC/6 (Aug.

30, 1996) (“[i]n any dispute, it is the burden of the maker of the database to dem-

onstrate the necessary investment”); see also Gaster, supra note 128, at

766 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

At the same time, users and competitors seeking to use

the sole-source data should remain free to challenge data-

base makers who opt for the second rate structure, and

should reap the benefits of a successful challenge by obtain-

ing the data at lesser of the statutory rate or the actual

cost.

345

Finally, to further encourage bargaining, and to fi-

nance the administrative mechanism to arbitrate actual cost

disputes, the loser could be assessed costs, including costs

incurred by the other party and administrative costs. Such a

dual rate structure could accommodate a negotiated “menu

of liability options”

346

analogous to Professors Reichman and

Samuelson’s proposal for a liability regime,

347

without un-

fairly saddling database makers with the bill for subsidizing

downstream data users.

3. Prohibiting Exclusive Licensing Arrangements

To further ensure that competition reigns in the data

market, exclusive licensing arrangements with sole-source

data producers could be deemed void against public policy.

Thus modified by introducing liability principles,

348

an in-

ternational sui generis regime modeled on the Database Di-

rective could ensure that competition performs its protective

role

349

even in the case of sole-source data.

350

345. If the database industry chose to form a collection society to minimize

transaction costs associated with licensing, the first appeal might lie within the

collection society. See supra note 342. Thereafter, either party could appeal the

actual cost determination to an appropriate court or, if the mechanism so pro-

vides, an arbitration committee.

346. Reichman & Samuelson, supra note 4, at 141.

347. Id.

348. See Initial Proposal, supra note 11, art. 8, O.J. C 156/03, at 9 (1992);

Amended Proposal, supra note 24, art. 11, O.J. C 308/01, at 13-14 (1993) (mandat-

ing compulsory licenses on “fair and nondiscriminatory terms” in the case of

sole-source data); cf. Reichman, supra note 10, at 2539-44 (proposing off-the-rack

liability regime with built-in compulsory license); see also Samuelson et al., supra

note 35, at 2412-14; cf. 17 U.S.C.A. § 115.

349. Zeidenberg, 86 F.3d at 1453 (citation omitted) (noting that competition

protects consumers in a free market economy).

350. See supra note 199 (defining “sole-source data”).

1997] EUROPEAN DATABASE DIRECTIVE 767

A compulsory licensing mechanism will not, however,

aid competition in markets without sufficient demand to

support additional competitors: neither will it aid competi-

tion in areas where information is provided under nonmar-

ket conditions, the conditions under which some scientists

perform basic research and development.

351

B. Potential Effects on Basic Research and Development

Assessing the role research

352

and development

353

(“R&D”) plays in technological innovation, and its resulting

societal effects, is an extremely complex task.

354

The com-

plexity of the task stems in part from the complexity in fund-

351. For example, scientists and other users can obtain publicly-funded in-

formation on a cost-recovery basis by the United States to scientists and other

users. See Freedom of Information Act, 5 U.S.C.A. § 552 (West Supp. 1996);

GLOBAL ACCESS, supra note 4, at 22; Reichman & Samuelson, supra note 4, at 113.

In addition, the Electronic Freedom of Information Act of 1996, Pub. L. No. 104-

231, 110 Stat. 3048 (effective October 2, 1996), requires federal agencies to pro-

vide that information in “any form or format requested . . . if the [information] is

readily reproducible by the agency in that form or format.” See 5 U.S.C.A. §

552(a)(3)(B).

352. Conventional innovation models, which presume a linear innovation

process, generally divide research into two levels: basic and applied. See

SCIENCE & ENGINEERING INDICATORS, supra note 23, at 4-9. “Basic research” seeks

to further scientific knowledge for its own sake, without regard to specific appli-

cations. Id. “Applied research” seeks to determine the means by which a par-

ticular, specified need can be met. Id. However, most innovation experts recog-

nize that the innovation process is far more complex than the linear model

suggests. See id. (citing Donald Stokes forthcoming book, P

ASTEUR’S QUADRANT).

Stokes classifies research into three levels: pure basic research (research con-

ducted solely in the quest for basic understanding), pure applied research (re-

search conducted solely for its potential use), and strategic research (research

conducted both for its potential use and for basic understanding). Id. As used

herein, “basic research” means pure basic research. In addition, except when

stated otherwise, this section examines data reflecting only United States re-

search and development patterns.

353.“Development” is the systematic use of the products of research,

knowledge and understanding, toward the production of useful products, proc-

esses, or services. SCIENCE & ENGINEERING INDICATORS, supra note 23, at 4-9.

354. The complexity stems not only from informational difficulties, but also

from differing analytical and philosophical approaches.

Id. at 8-3 (citations omit-

ted).

768 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

ing the R&D process,

355

from the difficulties in assessing the

spillover effects of R&D,

356

and from the difficulties in at-

taching value to the benefits derived from technological

growth.

357

This section does not attempt such a monumental

task, but rather attempts only a general assessment of the

potential effects of a sui generis regime which protects the in-

vestment of database makers, on a subcomponent of the

R&D process: “basic research.”

358

According to some of those charged with conducting ba-

sic research, even if an international sui generis regime is not

355. In the United States, R&D is performed and funded by industry, gov-

ernment, and academia. See id. at 4-2. Of the $171 billion spent on R&D in the

United States in 1995, 60% was spent on development, 23% was spent on applied

research, and 17% (approximately $29.6 billion) was spent on basic research. See

id. at 4-9 Although industry funds and performs the majority of overall R&D,

the majority of industry funds are spent on applied research and development.

See id. at 4-5 to 4-7 (reporting that, in 1995, industry provided approximately 59%

of total R&D expenditures in the United States, and performed approximately

70% of the total R&D). In the area of basic research, the federal government pro-

vides the majority of funding, while the academic sector performs most of the

basic research. SCIENCE & ENGINEERING INDICATORS, supra note 23, at 4-9 (report-

ing that federally-allocated funds comprised approximately 58% of the $29.6 bil-

lion spent on basic research); see id. (reporting that academia performed ap-

proximately 50% of basic research conducted in the United States in 1995).

Although all three sectors have experienced a reduction in the rate of growth in

dollars spent on research and development, academia is the only sector that has

experienced a constant dollar increase since 1986. Id.

356. Because basic research is often shared among scientists, e.g., by publi-

cation in scientific journals and by private circulation among colleagues, it is dif-

ficult for those conducting the research to capture the ripple effects, or spillover

benefits produced by the research. See id. at 8-6 to 8-7. In addition, initial re-

search may find application (“spillover”) in other industrial sectors and scientific

disciplines. See id. at 6-4, 8-6, to 8-7. Finally, the ripple effect of basic research

conducted in academic institutions represents a “crucial component in training

future scientists and engineers, many of whom will one day be working in [in-

dustrial] laboratories.” S

CIENCE & ENGINEERING INDICATORS, supra note 23, at 4-10.

357. For example, the difficulties inherent in attaching a monetary value to

technological breakthroughs that save human lives; or, in attaching a monetary

value to the more general benefit of promoting a “culture of reasoned discourse.”

See id. at 8-10 to 8-11.

358. Unlike “applied research,” which seeks to determine the means by

which a particular, specified need can be met, “basic research” seeks to further

scientific knowledge without regard to specific application. See id. at 4-9.

1997] EUROPEAN DATABASE DIRECTIVE 769

adopted, the Database Directive may cause a paradigmatic

shift in the way basic research is conducted.

359

This concern

rests on the argument that because the Database Directive’s

sui generis regime may create de facto property rights in basic

research data, the “principle of full and open exchange” will

give way to “mini-monopolies” which will in turn signifi-

cantly diminish long-term R&D capabilities.

360

The follow-

ing subsection delineates those concerns.

1. Concerns Expressed by Those Conducting Basic

Research

Opponents of the Database Directive’s sui generis regime,

and of an international sui generis regime modeled on the

Database Directive, cite three main concerns, all of which

center on the effect of such a regime on the accessibility of

data.

361

First, they argue that an international sui generis re-

gime will give database makers an absolute monopoly on

data.

362

Because of this exclusive right in data, critics claim,

the database maker will either deny access or charge a price

that science cannot pay.

363

359. See, e.g., GLOBAL ACCESS, supra note 4, at 27; Academy Letter, supra note

23, at 2 (discussing “ethos of sharing”); see generally THOMAS S. KUHN, THE

STRUCTURE OF SCIENTIFIC REVOLUTIONS (2d ed. 1970).

360. See GLOBAL ACCESS, supra note 4, at 14; Academy Letter, supra note 23,

at 2; Reichman & Samuelson, supra note 4, at 91; Rosler, supra note 11, at 140-41.

361. However, even those opposing the current proposals agree that the

“core idea underlying current sui generis proposals is sound, even if the mecha-

nisms proposed to address the problem are flawed.” See Reichman & Sam-

uelson, supra note 4, at 130. Additional concerns expressed by opponents were

addressed previously with respect to Database Directive’s effects within the EC.

See supra notes 277-321 and accompanying text.

362. See Academy Letter, supra note 23, at 1-2; Reichman & Samuelson, supra

note 4, 90-102 (arguing that each regime grants monopoly rights); G

LOBAL

ACCESS, supra note 4, at 35 (arguing that under “an exclusive rights model . . . a

database owner[] [has an] absolute monopoly”).

363. See Reichman & Samuelson, supra note 4, at 115-16; see also G

LOBAL

ACCESS, supra note 4, at 8-9. Although the sui generis regime may lead to higher

data costs, in the form of fees or independent generation costs, charging for data

is not a revolutionary concept. Information is not, nor has it ever been free—

someone, whether it be the government, or other information users who subsi-

770 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

Second, opponents argue that, because any entity can

find protection under such a sui generis regime,

364

foreign

governments will enter the database markets and charge

market prices for publicly-funded data previously supplied

on a cost-recovery basis.

365

Critics project that this develop-

ment, coupled with the incentive to privatize data, will re-

sult in an ever-shrinking pool of public domain informa-

tion.

366

Finally, opponents argue that the incentive to privatize

data created by such a regime is antithetical to the principle

of “full and open exchange” of scientific data.

367

Critics

speculate that the collective effect of this “parade of horri-

bles,”

368

which would follow adoption of such a sui generis

regime,

369

could have grave consequences to the United

States’ ability to maintain its lead in technological innova-

tion.

370

2. Potential Benefits to Those Conducting Basic

Research

Although de facto property rights in data causes much

concern to those whose lifeblood is data, a regime which

dize information pirates and parasites, has always paid the price. See Pamela

Samuelson, Information as Property: D. Ruckelshaus and Carpenter, Signal Changing

Direction in Intellectual Property Law?, 38 CATH. U. L. REV. 365, 367 (1989). More-

over, in some degree information has always existed as a commodity available to

those with the wherewithal to purchase it. Id.

364. However, note that H.R. 3531 specifically excludes governmental agen-

cies and employees in their official capacity from the scope of its protection. See

H.R.

REP. NO. 3531, supra note 1, § 2.

365. See Reichman & Samuelson, supra note 4, at 115-16; GLOBAL ACCESS, su-

pra note 4, at 15.

366. See Reichman & Samuelson, supra note 4, at 117, 128; see also GLOBAL

ACCESS, supra note 4, at 34-35.

367. See Academy Letter, supra note 23, at 2; Reichman & Samuelson, supra

note 4, at 114 (arguing that basic science requires an abundant infusion of “cheap

data in ways that encourage serendipity and playful exploration”).

368. See Reichman & Samuelson, supra note 4, at 128.

369. Id.

370. See id. at 115-17.

1997] EUROPEAN DATABASE DIRECTIVE 771

grants the commercial database maker a de facto property

right in data grants the same right to the dedicated-

researcher-turned-database-maker.

371

Thus, while the poten-

tial negative effects demand careful attention by policy mak-

ers before adopting an international sui generis regime mod-

eled on the Database Directive,

372

policy makers should

assess the potential negative effects in light of the potential

benefits to those decrying the Directive and the Proposals it

has spawned, and in light of the potential benefits to society

as a whole.

a. Market Participation

Those conducting basic research and their employers,

universities, for example, are not only data users, but are

also data producers. As data producers, researchers can en-

gage the market for protection. For example, the academic

and scientific communities might collectively organize and

form large databases fed continually by their respective

communities.

373

Given the sheer volume of information

371. Moreover, the term of the de facto monopoly will be determined by the

difficulty of unearthing the data and the relevant database market. That is, the

data collector will face price competition earlier, or data users will independently

generate the data, if data are relatively easy to independently discover and gen-

erate. Conversely, if data are relatively difficult to discover or generate, the data

collector may receive a longer term of protection.

372. These effects should also be considered by EC members when imple-

menting the Database Directive. See Reichman & Samuelson, supra note 4, at 108

(noting that even if an international sui generis regime is not adopted, the Data-

base Directive may force a paradigmatic shift in the way scientists, academics,

and others operating in a noncompetitive or subsidized environment, conduct

research and other information-intensive activities); see also G

LOBAL ACCESS, su-

pra note 4, at 25-26 (arguing that EC members should implement the Database

Directive by freighting “data generated by public funds . . . with a built-in, cost-

based discount for scientific research and education as a condition of further

commercialization by others”).

373. Although this may require some logistical nimbleness, given today’s

software tools, it takes relatively little programming expertise to form a database

which would find protection under the Database Directive. Moreover, the aca-

demic research community typically has ready access to programming expertise

and computer hardware. Griffith Interview, supra note 40. In addition, formation

772 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

produced by these communities,

374

an international sui gene-

ris regime might not only provide these communities with

significant bargaining power in the commercial sector, but

might also empower these communities to influence

fundamental information policy.

Moreover, armed with this new-found bargaining power,

the academic and scientific communities could engage the

market in a number of ways. For example, they could en-

gage commercial database makers in partnerships, provid-

ing data in exchange for innovative research tools and cost

recovery access.

375

Alternatively, the academic and scientific communities

could choose to enter the market as database makers.

376

As

market participants, these communities could employ the re-

sulting profits to replace ever-shrinking governmental

funds,

377

to pay for production costs, to pay for additional

research and educational activities, or to pay individual re-

searchers via a collection society.

378

of a collection society might reduce transaction costs which might otherwise

make small transactions cost-prohibitive. See Thomas K. Landry, Columbia-VLA

Journal of Law & The Arts Roundtable on Electronic Rights, 20 COLUM.-VLA J.L. &

ARTS 605, 645 (1996).

374. See supra note 320.

375. See Reichman & Samuelson, supra note 4, 108-10; cf. SCIENCE & EN-

GINEERING

INDICATORS, supra note 23, at 5-43 (noting that the strong upward trend

in gross revenues generated by academic patenting indicates an awareness and

willingness on the part of private industry to invest in academia) citing a 1994

survey conducted by the Association of University Technology Managers, Inc.

376. See Hunsucker, supra note 219, at 35-36; cf. S

CIENCE & ENGINEERING

INDICATORS, supra note 23, at 5-42 (noting that the sevenfold increase in the num-

ber of patents awarded to the academic sector between 1960 and 1980 suggests

that the academic sector is attempting to capture some of the economic benefits

associated with basic research).

377. See, e.g., id. at 4-21 (noting the decline, in real terms, of overall federal

R&D expenditures).

378. Cf. B

IERDERMAN, ET AL., LAW AND BUSINESS OF THE ENTERTAINMENT

INDUSTRIES 526-27 (3d ed. 1996) (describing ASCAP, BMI, and SESAC, collection

societies who collect and distribute fees generated by performing rights); see also

Landry, supra note 373, at 642 (describing the Authors Registry, a collection soci-

ety which collects royalties on electronic distribution of written articles).

1997] EUROPEAN DATABASE DIRECTIVE 773

b. An International Treaty Addressing Publicly-

Funded Data

While some governmental data are not reproducible and

may not be subject to the limits of a competitive market,

379

many of these data are presently available on a cost-recovery

basis pursuant to bilateral or multilateral treaties.

380

Addi-

tionally, although foreign governments may use the sui

generis regime to finance government-sponsored research ef-

forts, or may discriminate against researchers in the United

States,

381

there are practical limits on the ability of foreign

governments to do so.

Consider, for example, that the United States accounts for

approximately forty-four percent of the world’s R&D in-

vestment, outdistancing the next four largest investors com-

bined.

382

Governments who discriminate against researchers

in the United States, or who charge market prices for pub-

licly-funded data, risk similar treatment from the United

States. The threat of retaliation by the United States might

act as an implicit limit on the price foreign governments

charge for publicly-funded data.

Moreover, rather than face the wrath of its own citizens

by charging market prices for publicly-financed data, the

United States might exercise its strength in R&D to persuade

379. For example, data provided via governmentally-operated satellites.

See, e.g., GLOBAL ACCESS, supra note 4, at 22.

380. Id.; see also Bilateral Relations, Executive Summary 6 (visited Nov. 11,

1996) (noting that “during

FY-1995 there were 805 S&T [Science & Technology] agreements in force be-

tween the U.S. and 78 other countries.” For a representative listing of the current

bilateral agreements, see

In addition, one should note that international scientific col-

laboration and cooperation is increasing. See SCIENCE & ENGINEERING INDICATORS,

supra note 23, at 5-35-37 (noting greater international collaboration on scientific

and engineering articles); 4-42-48 (noting increasing international collaboration

in the public and private sectors).

381. See supra note 365.

382. S

CIENCE & ENGINEERING INDICATORS, supra note 23, at 4-35 and tables 4-

33-4-35 (note that these data reflect total R&D, not just basic research).

774 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

foreign governments to supply publicly-financed data to sci-

ence on reasonable terms.

383

Alternatively, foreign action

regarding publicly-funded data which the United States per-

ceives as unfair may precipitate a single international treaty

addressing publicly-funded data.

384

c. Reducing Inefficiencies Masked by Hidden

Subsidies

Subjecting basic research to the discipline of a commer-

cialized data market may reduce hidden costs imposed by

the systemic inefficiencies associated with research con-

ducted under less-than-market conditions. For example,

openly charging for data may give users a financial incentive

to make decisions based upon actual need, which may result

in more focused research.

In contrast, “free” access to information, which is not

really free but rather subsidized by someone else, may en-

courage unnecessary and unfocused research on systems

with finite capacities, which may render systems unavailable

to others. Thus, attaching a market price to information may

lead to wider data availability and accessibility by freeing

what would otherwise be unavailable system capacity.

d. Not a Threat to Technological Innovation

Even if imposing additional costs on basic research,

whether in the form of fees or of independent generation

costs, shifts the focus of basic research from the discovery of

knowledge for its own sake to the discovery of knowledge to

solve a particular societal problem, it does not necessarily

383. See, e.g., GLOBAL ACCESS, supra note 4, at 27; see also supra note 351 (cit-

ing the Freedom of Information Act).

384. For example, an international treaty might require governments to

supply their data on “fair and nondiscriminatory terms.” See supra note 338; cf.

GLOBAL ACCESS, supra note 4, at 25 (arguing that publicly-funded data should

come freighted with a built-in discount for science).

1997] EUROPEAN DATABASE DIRECTIVE 775

follow that such a shift will reduce the number of techno-

logical breakthroughs. On the contrary, some empirical evi-

dence suggests that technological breakthroughs are just as

likely to precede, rather than follow, basic research.

385

Thus, while the Database Directive’s sui generis regime,

or an international sui generis regime modeled on the Data-

base Directive, poses risks to academic and scientific re-

searchers, it also presents many opportunities to those same

researchers (or those making the substantial investment in

database production), including the opportunity to recover

some spillover subsidies they currently provide to the indus-

trial sector.

386

C. Potential Benefits to Society

Like the railroad of yesteryear, information is the engine

that drives commercial innovation and social advance-

ment;

387

it is the cornerstone of electronic commerce.

388

Mil-

385. See supra note 350 (regarding linear/nonlinear innovation models); see

also SCIENCE & ENGINEERING INDICATORS, supra note 23, at 4-9 (citing Donald

Stokes forthcoming book, PASTEUR’S QUADRANT, which sets forth examples of

real world technological breakthroughs accomplished by those such as Pasteur,

Faraday, and Kelvin, whose research focused on both fundamental understand-

ing and potential application; see also id. at 5-42 (noting that the academic sector,

which performs most of the basic research, is increasingly focusing on commer-

cially relevant technologies).

386. See supra note 354.

387. See WIPO Proposal, supra note 1, Preamble WIPO Doc. CRNR/DC/6

(Aug. 30, 1996) (“[D]atabases are a vital element in the development of a global

information infrastructure and an essential tool for promoting economic, cul-

tural, and technological advancement . . . .” ; 142 C

ONG. REC. E890-04 (daily ed.

May 23, 1996) (statement of Sen. Moorhead) (“[D]atabases . . . the linchpin of a

dynamic commercial industry in the United States.”); Database Directive, supra

note 1, recital 9, O.J. L 77/20, at 20 (1996); cf. Rockwell Graphic Sys. v. DEV In-

dus., 925 F.2d 174, 180 (7th Cir. 1991) (“The future of the nation depends in no

small part on the efficiency of industry, . . . [which in turn] depends in no small

part on the protection of intellectual property.”); see also Samuelson, supra note

363, at 367 (“In the information age, information becomes the primary economic

commodity, the source of greatest wealth”); see generally, Jared Sandberg, Micro-

soft Plans to Spend Big on the Internet, WALL ST. J., Nov. 15, 1996, at B3 (reporting

that Microsoft plans to spend $400 million this year to development information

content for its Internet services).

776 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

lions of information users, including scientists, professionals,

scholars, consumers, and businessmen and women, make

decisions based upon information every day.

389

To the ex-

tent that a sui generis regime protecting investment in data-

bases increases the production of databases, a tool that en-

hances information users’ ability to retrieve the kernel of

information needed to solve a particular scientific, legal,

economic, or medical problem, the regime would enhance

society’s problem-solving abilities, increase productivity and

decrease lost-opportunity costs,

390

advance education and

training, and facilitate the creation of a better-informed citi-

zenry.

391

Additionally, a sui generis regime protecting investment

in databases could produce profound long-term societal

benefits if such a regime subtly shifts macroeconomic incen-

tives from fictional works to more necessary factual works

392

having higher social utility.

393

That is, assuming that legal

protections create or protect incentives to produce, to the ex-

tent protection afforded factual works approaches or exceeds

the protection afforded fictional works, societal resources

388. See, e.g., Bradford L. Smith, Creating the Global Information Society: Look-

ing Ahead, in 2 FORDHAM INTERNATIONAL INTELLECTUAL PROPERTY LAW & POLICY

(Hugh C. Hansen ed., forthcoming 1997) (explaining that users do not want new

telecommunications equipment and computer hardware for its own sake, but

rather for the information it allows them to access)

389. 142 CONG. REC. E890-04 (daily ed. May 23, 1996) (statement of Sen.

Moorhead).

390. The lost opportunity costs consist of costs suffered by society in the

form of nonproductive time spent searching for information. See generally A.

MITCHELL POLINSKY, AN INTRODUCTION TO LAW AND ECONOMICS (2d ed. 1989).

391. See, e.g., 142 CONG. REC. E890-04 (daily ed. May 23, 1996) (statement of

Sen. Moorhead).

392. Although one could correct the imbalance by reducing the protection

currently afforded traditionally copyrightability works to little more than protec-

tion against wholesale copying, that solution seems unlikely given the vested in-

terests in the copyright regime. See generally Glynn S. Lunney, Jr., Reexamining

Copyrights’ Incentive Access Paradigm, 49 V

AND. L. REV. 483 (1996).

393. By according factual works only a thin veil of copyright protection be-

cause of their high degree of social utility, society has implicitly, if not explicitly,

deemed factual works more valuable than fictional works. See Lunney, supra

note 65, at 568-70.

1997] EUROPEAN DATABASE DIRECTIVE 777

should be redirected toward production of factual works.

394

Society would thus reap the benefit of increased production

of more necessary factual information.

CONCLUSION

Market-destructive appropriators, commercially-

motivated or not, skew incentives to invest in information

generation and compilation. The EC took a bold, intellectu-

ally-honest step forward by adopting a sui generis regime to

protect investment in databases. The Directive’s sui generis

regime does not create exclusive legal property rights in

data, but does protect the investment of labor and capital,

made by diligent compilers, in their compilations. By link-

ing protection to the investment of capital and labor, the Da-

tabase Directive’s sui generis regime should increase, or at

least preserve, incentives to collect and compile information

in the EC. Because independent creation and imitation re-

main, as they should, perfectly legitimate means of competi-

tion, the Directive’s sui generis regime protects compilers

without unnecessarily suspending the principles of free

competition in the nonproprietary data market.

At the same time, this Article recognizes the risk of im-

peding competition in the sole-source data market and the

incentives to privatize data that such a regime creates.

Nonetheless, there are sufficient ameliorative measures

available to those within the EC to offset these risks, not the

least of which is the fact that European courts will, if follow-

ing the spirit of the Directive, focus on the legitimate com-

mercial interests of the database maker when defining in-

fringement. On the other hand, because of the lack of similar

ameliorative measures at the international level, this Article

suggests that an international sui generis regime modeled on

394. See id. at 655-56; see Ginsburg, supra note 6, at 349 (the extent of copy-

right protection for databases influences the initial investment decision); see also

supra note 323.

778 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. [7:697

the Database Directive be modified by incorporating a com-

pulsory mechanism to protect the public’s interest in compe-

tition in the sole-source database and information markets.

However, this Article does not agree with those who be-

lieve information should be free. Society gains little by mak-

ing information free if, by so doing, it reduces the quantity

and quality of information produced, or reduces the mone-

tary value attached to socially-productive behavior like basic

research.

On the other hand, if increasing the price of data results

in greater remuneration to information producers, for exam-

ple, those engaged in basic research and science, society may

attach a higher value to such socially-productive behavior,

which in turn could result in more scientists and researchers,

who could in turn produce more information. Thus, greater

remuneration to researchers and scientists may result in

spillover effects that could have far-reaching social implica-

tions. What, for example, might children think if society

paid the next dedicated researcher laying the foundation for

the next Salk or Pasteur as much as the next Danielle Steele,

John Grisham, or Theodore Geisel?

395

395. Geisel is better known as Dr. Seuss; Steele and Grisham are successful

contemporary novelists.