Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Keywords: databases, copyrights, intellectual property, sui generis protection, European Union,
Feist v. Rural Telephone
ABSTRACT: This paper develops three arguments for increasing the strength of
database protection under U.S. law. First, stronger protections would encourage
private investment in database development, and private databases have many
potential benefits for science and industry. Second, stronger protections would
discourage extensive use of private licenses to protect databases and would allow for
greater public control over database laws and policies. Third, stronger database
protections in the U.S. would harmonize U.S. and E.U. laws and would thus enhance
international trade, commerce, and research. The U.S. should therefore follow the
European example and develop two tiers of protection for databases: 1) protection for
creative databases under copyright law; 2) protection for non-creative databases
through a special type of sui generis protection. In order to balance private control of
data and public access to data, sui generis protections should define a “fair use”
exemption that permits some unauthorized extraction of data for private, educational,
and research purposes, provided that such extraction does not adversely impact the
economic value of the database.
The Internet has created new forms of communication and has greatly enhanced
humanity’s ability to access and transmit information. A single person sitting down at
his or her personal computer can log on to millions of web sites around the globe and
Address for correspondence: David B. Resnik, J.D., Ph.D., Department of Medical Humanities, 2S-
17 Brody Building, The Brody School of Medicine, East Carolina University, Greenville, NC, 27858,
USA; email: resnikd@mail.ecu.edu.
Paper received, 5 March 2002: revised, 30 November 2002: accepted, 13 June 2003.
1353-3452 © 2003 Opragen Publications, POB 54, Guildford GU1 2YF, UK. http://www.opragen.co.uk
download gigabytes of data. The Internet has also had a significant impact on the pace
of scientific research and the strategies for conducting research. In the past, a
researcher would need to go to the library to browse through old journals to find a
particular article, a process that could take days, or even weeks, if the library did not
own the journal of interest. Today, he or she can find and download an article in less
than a minute using search engines and electronic databases. In the past, the period
between submission of an article for publication and its appearance in print could take
many months or several years. Today, submission, peer review, and publication can all
take place electronically, which decreases significantly the time between submission
and publication.1
Perhaps the most important effect that the Internet has had on scientific research is
that it now gives researchers access to vast quantities of data that they can search and
analyze. Indeed, the precursor to the Internet, the ARPANET, was originally developed
as a means of sharing data related to national defense research.2 For example,
biologists have available enormous quantities of DNA sequence data for many
different genomes, including the human, mouse, rice, corn, fruit fly, and Escherichia
coli genomes. They also have access to data on many different biologically significant
structures and compounds, including RNA, proteins, hormones, receptors, membranes,
cell types, tissues, and x-ray images.3,4 Using the analytical tools developed by the
emerging field of bioinformatics, biologists can now employ powerful computers to
search (or “mine”) these databases to discover correlations and patterns in the data and
a
to compare different databases.
This bountiful harvest of information comes at a price, however. It costs a great
deal of money to compile, maintain, service, and upgrade electronic databases. Unless
society decides to increase its support for public databases, private businesses will need
to foot the bill for many of these essential information services. In biomedicine, several
companies have developed private databases, including Celera Genomics, Monsanto,
a. For example, a scientist can discover a new type of cancer cell by comparing the cell to an
existing cell database. If the cell type is unique, he or she can sequence the DNA in the cell and
compare that DNA to DNA available in other databases to determine whether that cancer is
associated with a particular gene or mutation. If the scientist discovers a malfunctioning gene
associated with the cancer, he or she can use the sequence data from that gene to determine the
primary structure of the protein expressed by the gene, and he or she can use a protein database
to develop a model of that protein’s higher-order structures as well as its functions in cell.
Throughout all of his or her steps, the scientist will also benefit from searching publication
databases for data, methods, or tools relevant to his or her research, without access to these
enormous databases with powerful searching tools, the scientist would have been severely
hampered in his or her ability to discover the biochemical basis of this cancerous cell. Thus,
electronic databases can greatly increase the pace of research and efficiency of research. Data
mining can be an important method in scientific discovery when one has large amounts of data
but has not developed a hypothesis. One can search the data to discover patterns, which may
suggest a hypothesis. One may then confirm the hypothesis by obtaining additional data. Data
mining, which can be characterized as “data-driven” research, differs from the traditional
approach to scientific method, which can be characterized as “hypothesis-driven” research. A
number of authors have argued that there are epistemological and methodological problems with
data mining, but this paper will not explore these issues. For further discussion, see Glymour, C.
et al (1997). Statistical themes and lessons for data mining. Data Mining and Discovery 1:1-10.
b. “Sui generis” is Latin phrase used in legal and philosophical scholarship meaning roughly “self
made.” To say that something is sui generis means that it stands on its own or is not dependent
on other things. A database with sui generis protection has legal protection simply because it is a
database; its legal protection does not depend on its status as an original work under copyright
law or as an invention under patent law. There is a legal precedent for granting special sui
generis rights in intellectual property. The Semiconductor Chip Protection Act (17 U.S.C. 901)
provides special protections for computer chips that would not be available under copyright or
patent law. The U.S. Congress enacted this law in order to protect computer chips against piracy
and unauthorized copying and to provide legal protections for the semiconductor industry. See
Brooktree Corp. v. Advanced Micro Devices Inc., 977 F. 2d 1555 (Fed. Cir. 1992).
current options. Section 6 will present a brief overview of the strong database
protection laws adopted by the E.U. Section 7 will discuss briefly the implications of
international intellectual property treaties for database protection. Section 8 will
consider arguments for increasing the protection for databases under U.S. law. Section
9 will consider and evaluate different options for strengthening the U.S.’s database
protection laws, including two bills that the U.S. Congress debated. In evaluating these
different policies, Sections 8 and 9 will appeal to the moral and economic justifications
for intellectual property discussed in Section 3.
c. “Digital information” is simply information that has been converted to digital form. For example,
a photograph is not digital information until it has been scanned and converted into a digital
image. The significance of digital information is that it is very easy to store, transmit, arrange,
modify, analyze, and copy. This makes digital information very convenient to use but also very
difficult to control.
Intellectual properties are collections of rights to exercise dominion and control over
intangible things, such as poems, musical compositions, inventions, or computer
software. In contrast to tangible properties, intellectual properties are inherently non-
exclusive and non-rivalrous. That is, two people can both use the same intangible thing
at the same time without interfering with each other’s ability to use it. To gain
exclusive control over a toothbrush, one need only possess it; to gain exclusive control
over a poem, one requires some additional help from the government in order to
exclude other people from exercising dominion and control over the poem.7 Thus, laws
play an essential role in defining, limiting, and regulating intellectual property rights.
Copyright laws are rules for the protection of a type of intellectual property known as
an original work. Patent laws protect inventions, and trademark laws protect distinctive
features of commercial products and services. Trade secrecy laws protect confidential
business information.8
There are two basic justifications for intellectual property laws, the natural rights
approach and the utilitarian approach.9 According to the natural rights theory, all
human beings have some fundamental rights to life, liberty, and property.10 This theory
implies that human beings have rights to control their original works, inventions, and
other intellectual properties. The natural rights approach suffers from several
difficulties, however. First, one might argue that there are no natural rights; all rights
are derived from moral duties. Second, one might argue that the language of rights is
inherently divisive and counterproductive in political dialogue. To address important
policy issues, one should talk instead of obligations and duties instead of rights. To
address these critiques fully is beyond the scope of this essay,11 however, I would like
to explore a third critique of this view.
Assuming, for the sake of argument, that it makes some sense to talk about a right
to intellectual property, then one should consider whether that right is absolute and
inviolable, somewhat limited, or extensively limited. The primary function of rights is
to safeguard the person’s interests: rights are moral “trump cards” that one may use in
social debates to protect individuals. A right to liberty, for example, protects a person’s
interests in freedom of thought, movement, and association. If liberty is treated as a
basic right, then those who would restrict personal liberty must overcome a strong
burden of proof. One commonly accepted way of overcoming this burden of proof is to
argue that restrictions on personal liberty are required to prevent individuals from
harming other people. Another way to overcome this burden is to argue that some
restrictions on liberty are necessary in order to safeguard public health or safety.
If one uses the language of rights in social debates, it is important to recognize that
some rights are more important than others, since rights will often conflict. For
example, the right to life is generally regarded as more important than the right to
property. Suppose that a boy darts out onto the street while I am driving my car before
I have time to see him, and the only way I can avoid hitting him is to drive my car into
my neighbor’s fence. All would agree, I think, that the child’s right to life is more
important than my neighbor’s right to the enjoyment of his property. Suppose it is not
me, but a government official, who faces a similar dilemma concerning conflicting
rights. If there is a raging fire in the city that threatens thousands of houses and human
lives, most would say that firefighters would be justified in burning down someone’s
home if it is necessary to burn down the home to make a fire-break. The government is
justified in taking property in order to protect rights and promote important social
goals.
This same argument applies to intellectual properties as well: if there are natural
rights to intellectual property, they must be less important than most other natural
rights, including the right to life and liberty. The government may therefore sometimes
override, violate, or restrict these rights in order to safeguard other, more important
rights. Although this argument does not prove that there are no intellectual property
rights, it does prove that these rights can be sacrificed for a greater good, provided one
has convincing arguments for these restrictions. This argument also implies that most
of the real policy work in defining and delineating intellectual property rights resides in
the assessment of how these rights conflict with other rights and social goods. Thus,
society must strike some reasonable balance between protecting intellectual property
rights and promoting other values.8
One can reach similar conclusions from the utilitarian perspective on intellectual
property. According to the utilitarian view, the function of all rights is to promote the
maximum balance of good/bad consequences for society. Although rights do not
function as moral trump cards, they do draw attention to the importance of respecting
individual interests and allowing individuals to make their own choices.12 Individual
rights should be recognized and protected since the benefits to society of recognizing
those rights are likely to outweigh the harms. The utilitarian argument for recognizing
intellectual property rights is straightforward:
1. If society recognizes and protects intellectual property rights, this will provide
incentives to inventors and authors, entrepreneurs, and investors, which will
promote the development of science, technology, and the practical arts.
2. The benefits of the development of science, technology, and the practical arts for
society are likely to outweigh the harms.
3. Thus, society should recognize and protect intellectual property rights.
Article 1, Section 8, Clause 8 of the U.S. Constitution exemplifies this utilitarian
argument because it gives Congress the power “To promote the progress of science
and the useful arts, by securing for limited times to authors and inventors the
exclusive right to their respective writings and discoveries...(emphasis added).”13
In order to promote good consequences for society, it is necessary to strike the
appropriate balance between private and public control of intellectual property. If the
government grants too much control to private interests, then the public will lack
sufficient access to ideas and information. In order for progress to occur in science,
technology, or the practical arts, there must be a large public domain of ideas and
information that authors and inventors may use in creating original works and
inventions and in making discoveries. On the other hand, if the government does not
grant enough control to private interests, there will be insufficient incentives and
rewards for authors, inventors, entrepreneurs, and investors. Most of the intellectual
property laws in the U.S. and other industrialized nations have resulted from a careful
balancing of public and private interests.14 For instance, patent laws grant inventors
exclusive rights over their inventions, but they also require inventors to disclose their
invention to the public, and they limit the period of the patent to 20 years from the date
of application. Copyright laws give authors exclusive control over original works but
not over ideas and inventions.8 Indeed, the entire history of modern intellectual
property law, from British Letters Patent in the 15th century to current battles over the
control of the electronic information, embodies an ongoing dilemma concerning the
optimum balance of public vs. private control. These laws have had to change in
response to technological and social changes.
To summarize this section, regardless of whether one takes a natural rights or a
utilitarian approach to the moral and economic foundations of intellectual property
laws, the most fundamental policy question will be the same: how should society strike
the appropriate balance between private control of intellectual property and public
access to intellectual property?
the idea of a sheriff who wants to rid his town of criminals. One may copyright a paper
describing a mathematical proof, but one may not copyright the equations, formulas,
assumptions, or other parts of the proof. Spoken words are not copyrightable until they
are fixed in a tangible form, such as a written essay. Recalling the balancing idea
discussed in the Section 3, this aspect of copyright law reflects an important balance
between public and private control: if you write a book, then you may have control
over the copying of that book as it is expressed in tangible form, but you cannot obtain
control over the ideas and facts discussed and described in the book. Thus, someone
who reads a U.S. history book and extracts some of the facts that he read from the book
to write an essay does not have to obtain permission of the copyright holder to extract
those facts or use them in the essay. However, if he wanted to copy a substantial
portion of the book, he would need to obtain permission.
Someone who copies, performs, modifies, or displays a copyrighted work without
permission of the copyright holder infringes the copyright holder’s rights. In the U.S.,
copyright violations can lead to civil actions or a criminal prosecution. In a copyright
infringement lawsuit, the copyright holder (the plaintiff) must prove by direct or
circumstantial evidence that the defendant has infringed his or her copyrights. The
Patent Act defines an important exemption to copyright infringement known as “fair
use.” Someone who is sued for infringement may prove this defense to avoid legal
d
liability.
“Fair use” is not a clearly defined concept in copyright law. To determine whether
an infringement is fair use, the courts examine several factors, including:
• The purpose of the use (was it for commercial or non-commercial purposes?);
• The nature of the work (is the work a commercial or non-commercial work?);
• The proportion of the work used (was it a small part or a large part?);
• The significance of the part that was used (was it an important part or an
insignificant part?); and
• The economic impact of the use on the owner (did the copying lower the
commercial or economic value of the work?).
A court may find that copying meets the “fair use” exemption if it fulfills several of
these criteria, e.g. the use was for a non-commercial purpose, it constituted a small and
insignificant proportion of the work, etc.
In order to obtain copyright protection, the work must be original to the author. To
be original, the work need not be novel; the work need only be the tangible expression
of the author. What if the author does nothing more than make a copy or rearrangement
of an original work? Would this product have enough originality to be copyrightable?
For many years, the U.S. courts maintained the “sweat of the brow” doctrine for
determining whether a work is original: if the author has invested sufficient labor in
making the work, then the work is original to him. Under this doctrine, one may
copyright a poster made from an original painting, a rendition of a sculpture, or an
adaptation of a Beethoven sonata. As long as the variation between the original and the
d. The traditional view of “fair use” is that it is a defense to copyright infringement, although some
legal scholars interpret “fair use” as a positive right to make a copy.
copy is not merely trivial, then the person who makes a copy of an original has
copyrights on the copy.17
What if the original work is nothing more than a compilation of publicly available
facts, such as a database? Is a database copyrightable as an original work? These were
the questions the U.S. Supreme Court faced in Feist Publications v. Rural Telephone
Service Company in 1991.18 Feist Publications compiled and published area-wide
phone directories using data from local directories. Rural Telephone Service Company
was a local telephone company in Kansas, which had published a telephone directory it
had compiled from the data provided by its own subscribers. Feist sought to obtain a
license from Rural to use the data from its phonebook, but Rural refused to grant a
license. Feist then extracted the data from Rural’s phonebook to make its own
phonebook. Many of the listings in Feist’s phonebook were identical to those found in
Rural’s phonebook. Rural sued Feist for copyright infringement. The lower courts
ruled in favor of Rural, but Feist appealed the case all the way to the Supreme Court.
The Supreme Court reversed the lower decisions and held that Rural’s phonebook
was not copyrightable because it did not meet the requirements for a compilation under
the Copyright Act. To obtain copyright protection under the Copyright Act, the
compilation of facts must be selected, coordinated, or arranged in such a way that it
amounts to an original work of authorship. Rural’s compilation of facts did nothing
more than make an alphabetical arrangement of names, towns, and telephone numbers.
Although Rural invested some labor in making its phonebook, the phonebook was
completely obvious and lacked sufficient creativity. Thus, according to the Court’s
ruling in Feist, “the sweat of one’s brow” is not sufficient proof of originality. To
obtain a copyright, one must also demonstrate at least some minimal level of creativity
or novelty.8,19
Several cases involving West Publishing Company, a producer of legal databases,
also tested the extent of copyright protections for databases. Although West won a
couple of decisions in its home state of Minnesota, it lost an important case. In 1998,
the 2nd Circuit Court reaffirmed the Supreme Court’s holding in the Feist case in
Matthew Bender and Company v. West Publishing Company, ruling that West’s
compilations of publicly available legal cases, while exhibiting a great deal of
scholarship and labor, lacked sufficient creativity to constitute original works.20 The
11th Circuit also reaffirmed the ruling in Feist, holding that Warren Publishing
Company’s selection of data for its cable television “factbook” lacked sufficient
creativity and originality to obtain copyright protection.21
Like it or not, Feist is the law in the U.S. Moreover, since Feist addresses the
originality requirement for copyright, a statute that attempted to legislate against this
decision might be unconstitutional, because any such law would need to go against the
Supreme Court’s interpretation of the Constitution, but the Supreme Court, not
Congress, has the authority to interpret the Constitution.22 The implication of the Feist
decision is that copyright protection for databases in the U.S. is tenuous at best: since
protection depends on some finding of creativity by the court, the author may not know
in advance whether his database will be protected.
Moreover, even when a database can be copyrighted, this form of protection does
not extend to the facts contained in the database. Copyright protection for databases
only applies the original elements of the database. Thus, the owner of the copyrights on
a database cannot stop someone from extracting facts from the database to use in their
own database, provided that their database does not constitute an unauthorized copy or
derivative work.5
e. A “shrink wrap” license is a license contained on the software package that states terms and
conditions for using the software. A “click wrap” license is license agreement displayed on one’s
computer screen when downloading software from the internet.
for breach of contract because he is not a party to the contract. The licensor may be
able to sue the licensee if the breach results from his failure to perform his part of the
contract, however. For example, if A contracts with B to paint A’s house yellow, and,
unbeknownst to B, C paints part of the house purple, then A has no action in contract
against C. A may have a cause of action against B, if A has required B to take steps to
prevent others from painting the house a different color. Copyright law, on the other
hand, applies to many different people who may be directly or indirectly involved in
copying, distributing, displaying, or modifying original works. For example, a
television station could infringe a copyrighted broadcast by re-broadcasting an
unauthorized public display of the broadcast.8
Second, damages recoverable under a breach of a license agreement may not offer
the victim much of a remedy. In contract law, the victim of a breach is entitled to
recover damages sufficing to give him the benefit of the bargain.23 Thus, the law will
award him enough money to put him in the position he would have been in if the
contract had not been breached. In the case of a breach of a license agreement, it may
be difficult for the licensor to prove that he is entitled to extensive damages. If this is
the case, the court may award stipulated damages, if the contract includes such a
provision, or the court may award nominal damages. In contract law, punitive damages
are only awarded if the defendant is guilty of some type of misconduct, such as fraud
or oppression.23 If the damages are minimal, then the defendant will have little
incentive to settle out of court as well.
Third, it is impractical to enforce license agreements on thousands or even millions
of users who may not have “deep pockets”.8 A company would quickly go bankrupt as
a result of litigation costs if it had to sue for breach of contract to enforce its
proprietary interests in intellectual property. It makes far more sense for a company to
sue for breach of contract when it only needs to deal with a few defendants who would
be able to afford to pay for a breach of contract, or who would be motivated to settle
out of court.
Fourth, controlling intellectual property through licensing agreements presents
significant problems for public policy, since license agreements often expand
intellectual property protection beyond the scope of protection provided by copyright
or patent law. Since licenses are private agreements, the public has little input into
licensing practices, unless those practices are preempted by intellectual property laws
or violate anti-trust laws, fraud laws, or other laws. It is therefore more difficult for the
government to promote public access to intellectual property when use of the property
is governed by a private contract. On the other hand, the public has had a great deal to
say about intellectual property laws passed by Congress. If one wants to promote
public debate about the control of intellectual property as well as public access to
intellectual property, it is better to encourage legal reforms through changes in the
intellectual property regime rather than to rely on the generosity and good will of
privately contracting parties.24
In contrast to the U.S., European countries have strong protection for databases. In
1996, the E.U. approved a Directive on the Legal Protection of Databases.25 All of the
countries that participate in the E.U. must implement this Directive through their own
national laws. The Directive provides two tiers of database protection: the first layer
consists of the traditional copyright protection for databases; the second creates a new
sui generis form of protection for databases. The first tier protects all the original
aspects of the database; the second tier protects the data. The second tier prohibits
extraction or re-utilization of the whole or a substantial part of the database.25 This is a
much stronger form of protection for databases because copyrights do not prohibit
unauthorized extraction of facts. The sui generis form of database protection lasts for
15 years, and it may be renewed if the database owner makes a substantial change in
the database.
The Directive also creates numerous exemptions similar to the “fair use” doctrine
in copyright law. For example, it allows the extraction of data private use; for teaching
or scientific research, provided that proper credit is given and the use is for a non-
commercial purpose; and for the purpose of national security or the administration of
justice.25 A database does not need to be creative in order to have sui generis
protection: the Directive extends legal protection to databases where the creator has
invested substantial effort selecting, assembling, or verifying the database.25 Thus, this
form of protection bears some resemblance to the “sweat of the brow” doctrine that
governed U.S. copyright law prior to Feist.
Several international treaties that govern intellectual property have implications for
database protection. These treaties do not create international laws as such but allow
for national treatment of intellectual property.8 Every country that signs an intellectual
property treaty retains its sovereignty with respect to intellectual property: each country
still retains the ability to make and enforce its own laws. When countries sign an
international intellectual property treaty, they agree that they will bring their own laws
into conformity with the minimum standards set by the treaty. Signatories to the treaty
are still free to adopt protections stronger than the terms of the treaty, but those
stronger protections will have no effect on other signatories, unless those countries
agree to revise the treaty and adopt stronger laws.5
The Berne Convention, which was adopted in 1886 and has 130 signatory
countries, is an international copyright treaty administered by the World Intellectual
Property Organization (WIPO). Since the Berne convention protects literary and
artistic works, it would offer some international protection for creative databases, i.e.
databases that are copyrightable, but it would probably not protect non-creative
databases.5 The Berne Convention defines a minimum standard for the protection of
literary and artistic works. The Berne Convention does not apply to the E.U.’s sui
generis database protection laws because these laws provide protections much greater
than those found in the Berne Convention. Thus, even though the U.S. and E.U.
countries have signed the Berne Convention, the E.U. countries cannot require the U.S.
to enforce their sui generis database protection laws. In 1996 members of the WIPO
adopted a copyright treaty that clarifies the interpretation of international intellectual
property rights, but it does not offer any protection for non-creative works, such as
databases.5
The Agreement on Trade-Related Aspects of Intellectual Property (TRIPS),
adopted in 1995 as part of the General Agreement on Tariff and Trade (GATT), is
similar to the Berne Convention because it defines minimum standards for the
protection of intellectual property. The TRIPS agreement, like the Berne Convention,
protects creative databases but it does not protect non-creative databases.5 However,
TRIPS has several methods for enforcing intellectual property rights that go beyond the
policy of national treatment found under the Berne Convention. For example, if a
country violates the terms of the TRIPS agreement, a member of the World Trade
Organization (WTO) can request the formation of a panel to investigate and resolve
this dispute. It is possible that TRIPS may be amended at some point to include
protection for non-creative databases.5
To summarize, under current international agreements, the strong database
protections law adopted by the E.U. countries have no direct implications for database
protection in the U.S. Thus, someone who violates E.U. database protection laws in the
U.S. would not violate any U.S. laws and would not undermine the U.S.’s position in
international intellectual property treaties. However, a company that wanted strong
protection for a non-creative database could obtain that protection by conducting its
business in Europe. Thus, it is possible that database companies may leave the U.S. for
the Europe’s database-friendly legal environment.5
As described in Sections 4 and 5, the U.S. has some protections for databases under
copyright and contract. In addition, databases owners can use technological measures
to control access to databases. In thinking about database protection, it is useful to
recall the main conclusion from Section 3: intellectual property laws and policies
should seek to promote an optimal balance of public access vs. private control. Thus,
the key inquiry should be: do current U.S. laws and policies strike a fair balance
between these competing interests? There are at least three arguments for stronger
database protection.
many areas of research. Society can always develop laws that allow for public access
but also protect private interests. For example, society could adopt laws that create
“fair use” exemptions, as the E.U. has done.
protected under the E.U. legal regime. Eventually, E.U. countries would push for
changes in the TRIPS agreements, and the U.S. might resist such changes. None of
these developments would bode well for international database protection.
An opponent of stronger database protection could argue that there is no need to
harmonize U.S. and E.U. laws because countries should retain their own sovereignty
and do not need to adopt uniform laws. For example, capital punishment is legal in the
U.S. but illegal in Europe. If U.S. citizens decide that capital punishment is morally
justifiable, then the argument that there is a need to have international harmony with
respect to capital punishment laws should have little impact on the debate about capital
punishment in the U.S. Likewise, the E.U.’s stronger database protection laws should
have no effect on the debate about database protection in the U.S.
The problem with this argument is that a legal and moral issue like capital
punishment has little impact on international commerce, science, and industry, but an
issue like database protection can have a tremendous impact on international
commerce, science, and industry. Commonsense and basic economic theory both
suggest that international treaties and agreements with respect to issues that affect trade
and development are vital to promoting international commerce, science, and industry.
Even though countries retain their own national sovereignty, there are good reasons to
seek some degree of harmony concerning international laws that affect commerce,
science, and industry.
In the wake of the Feist decision, there has been a vigorous debate about whether the
U.S. should strengthen its laws for protecting databases. Many proposals for database
protection have been made, ranging from strong protections for databases to weaker
protections. In 2001-2002, two bills, HR 354 and HR 1858, were introduced in the
106th Congress, but neither bill became law. So far, no bills have been introduced in the
107th Congress, although the legislators remain interested in the topic. HR 354, the
Collection of Information Antipiracy Act, was modeled after the E.U.’s laws and
would have granted databases sui generis protection for a period of 15 years. Thus, HR
354 would have protected non-creative databases and the non-creative aspects of
creative databases. It would have also prohibited unauthorized extraction or use of a
substantial part of a database if the extraction or use harms the actual or potential
market of the database. HR 354 would have allowed unauthorized use or extraction for
research purposes unless the use or extraction damages the actual or potential market of
the database, and it would have allowed unauthorized use or extraction for educational
purposes unless the use or extraction is likely to be a market substitute for the
database.26,27
HR 1858, the Consumer and Investor Access to Information Act, would have
offered weaker protections for databases. Like HR 354, HR 1858 would have provided
databases with sui generis protection, but beyond that important similarity, the
proposed bill was very different. First, while the bill prohibited unauthorized copying,
it did not prohibit unauthorized extraction or use of data. The data in the database
would have remained in the public domain, even if the database itself were privately
controlled. Unauthorized extraction or use of data would have been prohibited only if it
amounted to a duplication of the database that resulted in unfair competition with the
original database. Thus, HR1858 would have granted all databases the type of
protection that is available to copyrighted works. It would have offered no protection to
the data itself.28
In the wake of the Feist decision, non-creative databases in the U.S. have no protection
under copyright law, although the law offers some protections for creative databases.
Moreover, copyright does not protect the data itself; it only protects the original work
that represents the data. Thus, if a private company in the U.S. develops a creative
database, someone could copy large portions of that database without violating
copyright law. If a company develops a non-creative database, someone could copy the
entire database without violating copyright law.
The arguments developed in Section 8 provide a sound rationale for strengthening
database protection in the U.S. First, stronger protections would encourage private
investment in database development, and private databases can offer many benefits to
researchers, companies, and consumers. Second, stronger protections would discourage
extensive use of private licenses to protect databases and would encourage greater
public control over database laws and policies. Third, stronger database protections in
the U.S. would harmonize U.S. and E.U. laws and thus enhance international
commerce, science, and industry. Thus, there is a very strong case that the U.S. should
do something to strengthen its database protection laws. But how much should the U.S.
increase its protection for databases? That is the key question.
To answer it, one should keep in mind that the main goal of intellectual property
law is to provide the best balance of public access to information vs. private control of
information. The E.U. Directive, which defines two types of protection for databases,
appears to strike a fair balance between public access and private control. Under this
law, a database could have copyright protection as well as sui generis protection. If a
database is sufficiently creative, then copyright law would protect the creative aspects
of the database, such as its selection or presentation of the data. In addition, the sui
generis protection would protect the data in the database for some minimum period of
time. In the E.U., sui generis protection lasts 15 years. This period seems to strike a fair
balance between public access and private control because it gives database owners a
period of time to obtain a return on their investment before allowing the data protected
in the database to become part of the public domain. If a database is non-creative, then
it still could have sui generis protection, provided that the creator of the database has
invested sufficient time and effort (“the sweat of one’s brow”) in developing the
database.
It is also important to define something like a “fair use” exemption in order to
strike a fair balance between public and private interests with respect to database
protection. Although some legal scholars and judges find the “fair use” exemption to
be somewhat frustrating and at times baffling, it gives the courts enough discretion and
leeway that they can develop rules that balance public and private interests. The “fair
use” exemption for sui generis protection of databases should be modeled after the
“fair use” exemption in copyright laws. Thus, it should permit some unauthorized
extraction of data for private, educational, and research purposes, provided that such
extraction does not adversely impact the economic value of the database.
REFERENCES
1. Lucky, R. (2000). The quickening pace of science communication. Science 289: 259-64.
2. Lemley, M. et al. (eds.). (2000). Software and Internet Law, Aspen Publishers, New York.
3. Roberts, R et al. (2001). Building a “genbank” of the published literature. Science 291: 2318-
2319.
4. Organization for Human Brain Mapping. (2001). Neuroimaging databases. Science 292: 1673-
1676.
5. Freno, M. (2001). Database protection: resolving the U.S. database dilemma with an eye toward
international protection. Cornell International Law Journal 34: 165-225.
6. American Library Association. (2002). Issue brief: database protection legislation.
www.ala.org/washoff/database.html (Accessed: February 18, 2002).
7. Conley, J., Brown, M., and Bryan, R. (2000). Database protection in a digital world: why the
United States should decline to follow the European model. Information and Communications
Technology Law 9, 1: 27-60.
8. Miller, A. and Davis, M. (2000). Intellectual Property, 3rd ed. West Publishing, St. Paul, MN.
9. Kuflik, A. (1989). Moral foundations of intellectual property rights, in: Weil, V. and Snapper, J.
(eds.) Owning Scientific and Technical Information, Rutgers University Press, Brunswick, NJ,
pp. 29-39.
10. Nozick, R. (1974). Anarchy, State and Utopia, Basic Books, New York.
11. Glendon, M. (1991). Rights Talk, The Free Press, New York.
12. Mill, J. (1956) [1859]. On Liberty, Liberal Arts Press, New York.
13. U.S. Constitution, Article 1, Sec. 8 (1789).
14. Resnik, D. (in press). A pluralistic account of intellectual property. The Journal of Business
Ethics.
15. 17 U.S.C. 101-512 (1976).
16. U.S. Copyright Office. (2000). Copyright Basics, U.S. Copyright Office, Washington, D.C.
17. Alfred Bell & Co. v. Catalda Fine Arts, Inc. 191 F. 2d 99 (2nd Cir. 1951), 298.
18. Feist Publication v. Rural Telephone Service Company. 499 U.S. 340 (S.Ct. 1991).
19. Bitlaw. (2002) Database legal protection. www.bitlaw.com/copyright/database.html (Accessed:
February 18, 2002).
20. Matthew Bender and Company v. West Publishing Company. 158 F. 3rd 693 (2nd Circ. 1998).
21. Warren Publishing v. Microdos Data Corp., 115 F. 3d 1509 (11th Cir. 1997).
22. Ginsburg, J. (1992). No “sweat”? Copyright and other protection of works of information after
Feist v. Rural Telephone. Columbia Law Review 92: 338-348.
23. Rohwer, C. and Skrocki, A. (2000). Contracts, 5th ed, West Publishing, St. Paul, MN.
24. National Academy of Science. (2000). The Digital Dilemma, National Academy Press,
Washington, DC.
25. European Union. (1996). Council Directive No. 96/9, 1996 O.J. (L 77/20).
26. Gardener, W. and Rosenbaum, J. (1998). Database protection and access to information. Science
281: 786-87.
27. Mauer, S. and Scotchmer, S. (1999). Database protection: is it broken and should we fix it?
Science 284: 1129-1130.
28. Keck, R. and Goode, D. (2001). Of misappropriated manure heaps, rude robots, and broken
promises: the (d)evolving law of database protection. The Business Lawyer 57, 1: 513-540.