DATABASE PROTECTION IN EUROPE – RECENT DEVELOPMENTS AND A MODEST PROPOSAL

The controversial provisions in the European Union’s Database Directive have created considerable uncertainty for commercial producers of databases, while recent case law has emasculated much of the Directive. However, researchers and academics must still work in a restrictive copyright environment within Europe. This paper reviews the Directive in the light of two recent UK reports that suggest a more liberal copyright regime is both culturally and economically desirable. The author suggests that unfair competition problems should be addressed by new unfair competition laws for Ireland and the UK and not through revision of the Directive.


INTRODUCTION
The 1996 Database Directive 1 was initially sold to Europeans as an innovative measure, one that would stimulate the European database industry, an industry that was seen even by the European Commission as lagging behind other competitors, particularly the USA 2 .The twin track approach to protection, providing a "harmonised" copyright for the makers of an original database, while at the same time providing a separate sui generis (of its own kind) right in respect of the unauthorised extraction and/or reutilisation of the contents of the database, provided a measure of cohesion for the European information industry.However, it did not take long for flaws in the Directive to emerge 3 .These include: vague and at times confusing drafting and language 4 ; confusion about where the boundary between the two rights lies 5 ; uncertainty over the duration of the sui generis right 6 ; 1 Adopted on 11 March 1996; 96/9/EC; OJL77, 27.3.1996.For an early critique see Hicks, Copyright and Computer Databases [1993] EIPR 113. 2 See the Explanatory Memorandum to the first draft of the proposed directive -COM (92) 24 final -SYN393 Brussels, 13 May 1992, paragraphs 1.2 and 2.1.1. 3See generally the discussion in Davison, Mark J., The Legal Protection of Databases (Cambridge) 2003 Ch. 3 (hereafter Davison).For a U.S. perspective, see Reichman, J.H. and Uhlir, P.F., Database Protection at the Crossroads [1999], 14 Berkeley Technology Law Journal, available at http://www.law.berkeley.edu/journals/btlj/articles/vol14/Reichman/html/reader.html. 4E.g., the (over)broad definition of database (Davison, p. 71) and the meaning of "obtaining, verification and presentation" in article 7, problems not entirely dispelled by the William Hill/Fixtures Marketing judgements: see Davison  Data Science Journal, Volume 6, Open Data Issue, 17 June 2007 of the collection could be protected from copying by reference to a substantial copying test.Again, cases indicated that protection would be available, but that facts could not be protected per se 14 .Certain defences, such as research and private study, criticism, or review would also prove to be useful to someone building upon earlier collections 15 .
Certainly, there were a few cases where database creation was inhibited in English law, the rights being invoked by commercial publishers against pirates who had borrowed or plagiarised the contents of commercial collections 16 .But there were also instances in which fact-based works (e.g., maps) were facilitated 17 .The Directive changed all that for the British and Irish.Those countries surrendered their low threshold originality test and broader exemptions from liability, especially for content extraction and reutilisation, obtaining in return: a higher threshold originality test for the compilation (author's own intellectual creation), a right that only protected a compilation qua compilation; and an uncertain sui generis right dependent upon an equally uncertain substantial investment test with limited defences confined to the lawful use and extraction of contents only.Re-use was prohibited, making allegations that the law was over-protective difficult to refute.
Other EU jurisdictions did not have as much to accommodate when transposing the Directive: the Scandinavians 18 thought the sui generis right was just an extension in the existing Catalogue Rule 19 ; the originality test was assumed to be only a restatement of the (high) 'personality of the author' test found in many Continental codes 20 ; earlier decisions 21 against protecting T.V. listings, phone books could be accommodated, especially in the Netherlands, via "spin-off" theory etc.; and many continental countries have narrower defences and exemptions than the UK and Ireland and most have levies for use, whereas Ireland and the UK are opposed to levies 22 .
National courts in the EU generally interpreted the copyright provisions narrowly.National courts also interpreted the sui generis right narrowly, either by saying the collection in question was purely factual or had no significant separate existence, being merely a 'spin-off' from a core business (e.g., T.V. listings, phone books).The suspicion exists that many national courts did not consider whether the author's own intellectual in plaintiff's compilation) when re-arranged and supplemented by defendant's notes on Hazlitt, of a kind inferior to plaintiff's own notes on the author. 14Cramp v. Smythson [1944] AC 329; Warwick Film Productions Ltd. v. Eisinger [1969] 1 Ch.508.creation test in the Directive was lower than the national originality standard, e.g., for literary works, but because both the compilation copyright and sui generis right were seen as being intellectual property (I.P.) rights (rather than as an I.P. right and an unfair competition law respectively), the Directive was doomed at the outset.

THE EUROPEAN COURT OF JUSTICE'S NARROW INTERPRETATION OF THE SUI GENERIS RIGHT
In a set of four cases 23 , the European Court of Justice (ECJ) ruled on the meaning of "substantial investment", for the purposes of a database holder being able to invoke the sui generis right against persons who had used database contents without permission.The Court drew a distinction between the work and investment that goes into the capture and of data (which is not protected), and the work and investment needed to obtain, verify and present the data (which is protected).Thus, "back office" expenditure and work is problematical, but "front office" or end-user related investment is protectable.This distinction greatly reduced the scope of the sui generis right.It is certainly not representative of UK and Irish pre-Directive case-law 24 and it is also not a meaningful distinction in US case-law under Feist 25 .
The consequences of the ECJ's William Hill decision for the UK and Ireland are extreme, at least for database producers.Neither of the binary rights outlined in the Directive will be available in many instances.Given that neither the UK nor the Irish have a coherent unfair competition law, commercial predators are able to utilise low-brow, investment-rich databases.As the compilation copyright may not be available at all, because the database may be a low-authorship collection of materials that do not constitute copyright able material, the author's own intellectual creation test may be set too high.Even if the test is not set too high, substantial taking from a database may be difficult to determine under English law 26 , should copyright compilation infringement remain open to a plaintiff as some commentators, including Davison, appear to think 27 .After William Hill the possible use of the sui generis right appears to be highly unlikely if the investment and effort goes into creation rather than presentation of the works or materials28 .The prospects of using passing off are not bright according to the relevant case-law29 .
For the scientific and academic communities, William Hill represents a good decision insofar as it suggests that the sui generis right is now virtually dead in the water.Perhaps this is not a balanced view, however.Some collections are commercially valuable and their legal I.P. protection can be justified in many instances.The real issue is how to reconfigure national laws in such a way as to provide a more balanced approach to the rights of ownership, both in terms of private entitlement to compilations of data and the broad public interest.
It may be that the way forward is to opt for a more pragmatic, case-by-case analysis based upon unfair competition principles rather than by the construction of a new neighbouring exclusive property right.In this context it should be possible to draw distinctions between commercial and non-commercial uses, and research and transformative uses of databases, as well as injecting flexible mechanisms that reflect national concerns and local cultural traditions.Indeed, one suspects that several European jurisdictions have interpreted both the database copyright and the sui generis right strictly simply because domestic competition law and consumer protection law often provide more intelligible misuse standards that those afforded under the Database Directive.
The paradox now is that US database protection may be more extensive than it is in Europe.Compare, for instance, the Dutch case, NVM v ZAH 30 (2006), with the US case, Montgomery County of Realtors v Realty Photo Master Corp. 31 (1995).
The facts of these cases are similar.The plaintiff produced an on-line real property sales website carrying descriptions of the property and photographs.The defendants reproduced this material on their own sites, appropriating the investment made by the plaintiff.The US Court, applying Feist, found the contents and the database were protected as literary and artistic works and as a compilation.On the question whether individual items describing the location of the property and describing the property itself could be protected, the US Court concluded that the descriptive material was "marketing puffery that cannot be described as factual" 32 -in other words, literary works in themselves -while the Dutch Court simply held that the descriptions were purely factual and not protected.
The Dutch Court found insufficient literary or artistic merit in the descriptors and photographs (for copyright purposes) and that the investment made was insufficient to attract the sui generis database right.The US Court specifically said that the earlier Feist decision did not deny copyright protection to the plaintiff.

Feist does not help [defendant's argument]. RPM argues that MCAR's database, like the telephone book in Feist
, is merely a collection of simple facts entitled to no copyright protection.Unlike a telephone directory however, the arrangement of the information in MCAR's database, "possesses at least some minimal degree of creativity." 33

THE EUROPEAN COMMISSION'S REVIEW OF THE IMPLEMENTATION OF THE E.U. DATABASE DIRECTIVE
In late 2005, the European Commission reviewed the Database Directive 34 and outlined for options for the European Union: Option 1 Repeal the Directive and allow all previous forms of protection to be reintroduced.

Option 2
Withdraw the sui generis right and continue with the unified copyright standard.

Option 3
Amend the sui generis right to bring the cost of creating the data to come within the substantial investment standard.

Option 4
Preserve the status-quo and do nothing.The European Commission's on-line survey of European database producers came up with the figure that 80% of respondents felt "protected" or "well protected" by the sui generis right, while in the separate survey of those database producers 36% felt the ECJ rulings had "weakened" or "eliminated" the sui generis right.
Following publication of the first evaluation Report 35 , the European Commission invited responses from interested parties.Fifty-five responses were obtained, the majority of which were from database producers.Options 3 and 4 were the most popular choices, hardly surprising given the nature of both the exercise and the identity of the contributors.Interestingly enough, a number of US voices were raised against the sui generis right and its continued existence.
Jonathan Band and others (Band, et al.) commented 36 that "the sui generis right should be withdrawn".They argued that the European Commission itself acknowledges that: (i) the right has not contributed to improved production figures of databases 37 ; (ii) its scope is difficult to determine and created difficulties of interpretation 38 ; and (iii) it has created confusion among certain users, in particular the academic and scientific communities.
Band Other sectors have taken a different perspective, however.For example, the Data Publishers Association (DPA) rejects the Gale statistics, which track the relative decline of the European database industry.The DPA suggests both that their own survey shows that the European database industry is expanding and that the European Commission needs to undertake more statistical work on the economics of the European database market.The DPA thus argues in favour of retaining of the legal status quo 39 .
All European Academies (ALLEA) took the opportunity to reiterate their position taken earlier on the Directive.The copyright provisions are regarded as operating satisfactorily by ALLEA because of the exclusion of mere facts or data from the provisions of copyright and because of the scope of the (copyright) exceptions.However, the sui generis right is criticised because the exceptions: are more tightly drawn than in the case of copyright law; ALLEA also indicated that a preference for Option 1 or 2 was favoured, but stated that if the sui generis right were to survive, ALLEA "would press for a modification of the Exceptions ….. to make them at least as extensive as those traditionally enjoyed under copyright" 40 .
The British Academy Review Group on Copyright 41 expressed a number of concerns.Their report stated that "the Database Directive is at once vague and wide-ranging and fails to contain the exemptions for private study and research and for criticism and review, both in respect of database copyright and the sui generis database right.These exemptions are essential to the balance of rights and interests inherent in copyright and associated legislation.As an increasing proportion of material takes a digital form, the scope of even the current exemptions may be seriously undermined".
The Review Group in particular highlighted the narrow defence of non-commercial teaching or research by a lawful user, a defence that only applies to acts of extraction.No defence in relation to reutilisation exists and no "criticism or review" defence is available, even to a lawful user.The Review Group also pointed to the oppressive commercial clout that compilers of databases may have when producing databases from single sources or through considerable financial "muscle" that can lock out competition and create "what is virtually a natural monopoly".In recommendation 9, the Review Group indicated that these trends needed to be monitored by public authorities and academic communities so as to ensure that the assertion of database rights and the changes made in the law to access to database contents do not become a growing impediment to scholarship.
On the critical question of the research and private study exceptions the Review Group, in recommendation 3, made a number of very pertinent observations that could, it is suggested, form the basis of either judicially adopted benchmarks or statutory presumptions: Under its wording, "research" is to be treated as distinct from "private study"."Research" for these purposes should not only encompass the initial stages of an academic project when material is being collected, but also subsequent stages, which involve the analysis and publications of the results.
"Research" in this context should be regarded as "non-commercial" in any circumstances where the taking of copyright material is fair and the presentation of the results will be without charge to the recipients, or will be at a price that can only be expected to cover the reasonable costs of production and distribution, including the reasonable profits of a commercial publisher.
Research that is financed by a research council or charitable foundation is presumptively noncommercial.
Charges that are not covered by the exemptions, because the research to which they relate is commercial, should be reasonable and competition authorities and the copyright tribunal should be able to restrain abuse.
One could go further by recommending that the propositions found in recommendation 3 should be adopted as statutory presumptions, or at the very least, statutory definitions that could be inserted into copyright law generally, requiring any argument seeking to displace the presumption or statutory definition to be made in copyright litigation by a person or legal entity that holds a contrary view."Soft law" provisions of this kind 40 Ibid.

THE GOWERS REVIEW OF INTELLECTUAL PROPERTY
Most recently, the UK Chancellor of the Exchequer, Gordon Brown -widely regarded as Tony Blair's likely successor as Prime Minister in 2007 -commissioned a report on the state of the intellectual property system within the United Kingdom.A primary task was an evaluation of those laws to determine whether they deliver incentives, while minimising inefficiency 43 .The Review, produced by Andrew Gowers and published on December 6, 2006 mentions database protection only in passing 44 , but, significantly, the review remarks that "new technologies such as….databasesrequire IP protection but do not fit easily into existing categories" 45 .Although it is not clear that Gowers is actually criticising existing database protection legislation, there are other parts of the Review that suggest that greater flexibility may be injected into UK intellectual property law in general.
A number of recommendations, therefore, are very pertinent to the database debate.Gowers comments that fair dealing and other defences are much narrower than the US fair use defence and that "this is stunting new creators from producing work and generating new value" 46 .This general observation finds specific expression in the form of recommendation 11, which suggests that the Information Society Directive should be amended to allow for "creative, transformative or derivative works" to be possible without infringing copyright, as long as the work in question can subsist within the parameters of the Berne Convention three-step test in article 9. 47 Gowers also seeks a broadening of the research exemptions in existing UK copyright law so as to "help research institutes to further knowledge by using ideas protected by others" 48 , but Gowers is rather weak and unclear about how to achieve this 49 , in contrast with the British Academy Report discussed above.
Gowers is rather more helpful in relation to anti-competitive practices, opining that competition authorities must have the ability to curb any abuse or monopoly power stemming from IP rights 50 .However, the Review again is not very specific on how such an outcome is to be achieved, pointing out that the Unfair Commercial Practices Directive 51 is not likely to be of much use in a business-to-business transaction, or one in which a consumer is not involved.

CONCLUSION
The ambitious programme of copyright harmonisation that started in 1988 may have effectively run its course.One does not detect any real enthusiasm within the European Commission for any new legislative process that would be as controversial or as acrimonious as that which surrounded the Information Society Directive, to name just one example.Indeed, the work in hand, on collecting societies and the enhancement of criminal and Hugenholtz, Football Fixtures, horse races and spin-offs: the E.C.J. domesticates the database right [2005] EIPR 113. 5 Davison at p.81; Kay, The Proposed EU Directive for the Legal Protection of Databases [1995] EIPR 583. 6Smith, Legal Protection of Factual Compilations and Databases in England -how will the Database Directive change the law in this area?[1997] IPQ450; Chalton, The Copyright and Rights in Databases Regulations 1997 [1998] EIPR 178.
, et al. see the Directive as an illustration of the dangers of over-protection and that a shrinking public domain can stifle innovation.While the Directive recognised that I.P. rights in facts are not the traditional bailiwick of copyright laws, the Directive may come precariously close to protecting basic information, as the European Commission 2005 Report itself acknowledges.Band, et al. see the decline in the European database industry as being a result of this (over) protection.They also argue that the repeal of the sui generis right would stimulate the expansion of the publishing sector because producers who 'buy in' content may still invoke the sui generis right and lock up information from competitors.
The contribution of Band, et al. is also directed at undermining demands from publishers in the United States for a similar sui generis right.The publishers in the United States have argued that the reciprocity provisions mean that U.S. databases in Europe are not protected because of the absence of a U.S. counterpart law.
to teaching and research vis-à-vis extraction only and not to re-use by teachers or researchers; unclear in scope and meaning; and are incapable of applying to any commercial outputs from research which is unacceptable and impractical: the authors argue that research outputs are difficult to quantify and difficult to disentangle from research activities per se.
38Jonathan Band's views on the state of the debate in the USA can be gleaned from writings such as "Response to the Coalition against Database Privacy Memorandum" (2004),21The Computer and Internet Lawyer 7 and "The Database Debate in the 108 Congress: Déjà Vu All Over Again" (2005),22The Computer and Internet Lawyer 1. 39 www.forum.eurpoa.eu.int/Public/irc/markt/markt_Consultations/library.are 42t in any way confront the existing Community acquis, that is, the rules and principles that can be built up in European legislation, because such expressions are not fully defined in the Copyright Directive.42 41Copyright and Research in the Humanities and Social Sciences (September 2006).(A working group of 8 members, which included Professor W.R. Cornish and Professor H. MacQueen), available at www.britac.ac.uk/reports/copyright/contents.html, as the Kay Report, named after the Chairperson, Professor John Kay. would