Out-Law / Your Daily Need-To-Know

The first case to look at database right and to consider the meaning of some of its expressions.  An important case showing the relevance of the new right for streaming data and the scope of the protection it provides.

British Horseracing Board, Jockey Club, Weatherbys v William Hill Organisation Ltd

  • [2001] 2 C.M.L.R. 12
  • [2002] E.C.C. 23
  • [2001] E.C.D.R. 20

Help with citations


The British Horseracing Board ('BHB') is the governing authority for the British horseracing industry and carries out a large number of administrative functions.  These functions reflect the size of the racing industry in the UK: for example, there were, in 2000, 1,209 race programmes and fixtures, 59 racecourses, and race meetings were held on 327 days of the year.  There are some 15,000 horses in training, 9,000 active owners and 1,000 trainers.  Each owner has unique colours in which his horses run.

BHB spends some 4 to 5 months each year compiling, granting and publishing the Fixture List.  Publication of this in July of the preceding year marks the start of the full process which culminates in weekly advertisements in the Racing Calendar of the races (for which entries are taken centrally under the Orders and Rules of Racing).  There are approximately 7,800 races run each year, and in 2000 there were some 175,000 entries for races.  Not all horses run, since some may be withdrawn or the race may be oversubscribed.  Shortly before each race, trainers must make a "declaration" of the horses which will actually run in a particular race.  In 2000, there were some 80,000 declarations of this nature, and 80,000 related declarations of riders.

BHB maintains a computerised collection of information (the "BHB Database") which is constantly updated with the latest information.  This involves extensive work, collecting raw data, designing the database, the selection and verification of the data for inclusion and the insertion and arrangement of selected data in the database.  The cost of obtaining, verifying and presenting its contents is approximately £4 million per annum and involves some 80 employees, as well as much computer hardware and software.

As the database has been in existence for some time, the BHB Database contains a vast amount of relevant information about owners, trainers, horses and much else.  It contains "pre-race information", which is made available in advance of a race, covering place and date of a race, distance, criteria for entering the race, date for entries to be made, entry fee, and so on.  Nearer the date of the race, the information is expanded and includes the time at which the race is scheduled to be run, and the list of the horses entered.  The final stage of pre-race information includes the "declared" runners, riders, weight, starting stall and owners' colours.  All this information is taken from BHB's computerised database.

Data is extensively checked, including checking details by telephone, with recordings made to allow independent verification of the details by someone other than the person who made the telephone call.

BHB is self-financing, and obtains its income from fees for registrations and licences, fixture fees from racecourses and entry handling charges payable by owners and racecourses.  Part of its income derives from fees charged to third parties for use of information contained in the BHB Database.  This source provides income of around £1 million per annum.

Information from the database is supplied to a wide variety of people, including elements of the racing industry (such as racecourses and owners); by the website www.bhb-weatherbys.racingadmin.co.uk, by the Prestel network, and also every week in the Racing Calendar.  It is used by journalists and broadcasters as well as other information services (such as Ceefax).

Information is supplied also to bookmakers by two routes.  Firstly, data is supplied to Racing Pages Ltd, controlled and owned jointly by Weatherbys and the Press Association.  Racing Pages makes available data to subscribers via a datafeed known as the "Declarations Feed".  This is an accurate and up-to-date list of races, declared runners and jockeys, distance, race times, number of runners and other information.  Secondly, data is supplied to Satellite Information Services Ltd ("SIS") which is allowed to forward data to its own subscribers in the form of a raw data feed (“RDF”).

William Hill is the nationally known bookmakers.  It does business by Licensed Betting Offices ("LBOs") and also by telephone betting.  At the LBOs, information is provided by means of data from SIS which gives live coverage by television and also a text service provided by William Hill on television screens.  Screens are updated during the day with the latest information as to odds and results.  Pages from the Racing Post are also displayed on the walls, and these provide much more information than is possible with the screens.

William Hill subscribes to SIS audio and video service, and has its own promotional audio material.  William Hill provides its own text service to its LBOs via a satellite channel provided by SIS.  While much of the information displayed derives ultimately from the BHB Database, no objection was taken to this, and BHB considered that these various uses were licensed and that it was being adequately recompensed for this.

The dispute related to the recent commencement by William Hill of betting services offered over the internet.  The public could access these webpages, see the horses that were running, when and where, and the odds offered by William Hill.  They could place bets by the internet as well.

William Hill provided internet betting on selected races between May 1999 and February 2000, but after that started to offer betting on all mainstream racing in the UK.  The data for the races was derived from the RDF taken from SIS.  However, by the time William Hill published this data, it was available from several sources other than SIS, for example, the television and the press as well as on several teletext services.  Thus, William Hill took its data from SIS, which SIS in turn had received from data on BHB's computer.  It was information which was also to be found in the Declaration Feed made available through the Racing Pages.  SIS had no right to sub-license William Hill to use any of BHB's data on its internet site, and had not purported to do so.  However, the data in question was data identical to information supplied to SIS from the BHB Database.

BHB claimed that it owned the database right in the BHB Database and that William Hill was making unlicensed use of it in its internet business.  BHB said that William Hill took data from the SIS RDF, which was itself derived from the BHB Database and this infringed the database right in two ways.  Firstly, each day's use by William Hill of the data from the SIS RDF was an extraction or re-utilisation of a substantial part of the contents of the database contrary to Article 7(1) of the Directive.  Secondly, even if the extracts were not substantial, the totality of William Hill's actions amounted to repeated and systematic extraction or re-utilisation of insubstantial parts of the contents of the database contrary to Article 7(5).

William Hill in reply accepted that the database at any given time qualified as a collection of data in which database right existed.  However, the database was dynamic and constantly changing, and so there was a point as to whether it was the same database for the purposes of the law as the collection of data even a few days later. Furthermore, there are alternative sources for the data which William Hill could use and which would avoid any allegation of infringement.


Database right was introduced by a Directive.  It was intended to harmonise the law among the Member States by introducing some harmonisation in copyright law as well as introducing a new right, described as sui generis.  This description was important since, although there are some similarities between copyright law and database right, the scope of the new right had to be determined from the Directive itself and not by applying copyright principles by analogy.

The definition of "database" in the Directive at Article 1 was important:

“This Directive concerns the legal protection of databases in any form.

For the purposes of this Directive, "database" shall mean a collection of independent works, data or other materials arranged in a systematic or methodical way and individually accessible by electronic or other means.

Protection under this Directive shall apply to computer programs used in the making or operation of databases accessible by electronic means."

The Recitals to the Directive expand on this definition.  Thus, by Recital 17, a recording, or an audio visual, cinematographic, literary or musical work as such does not fall within the scope of the Directive.  Further, by Recitals 15 and 16, it is expressed that originality (in the sense of the author's own intellectual creation) is the sole criterion for copyright protection of the database, without the necessity to prove any aesthetic or qualitative criteria.  However, Recital 21 makes it clear that for the database right to exist, the requirement that works, data or other material should be arranged systematically or methodically does not mean that those materials should have been physically stored in an organised manner.  Thus, "database" has a very wide meaning, covering virtually all collections of data in searchable form.

Understanding the purpose behind database right is important as it throws light on what types of activity amount to breach.  Based on the Directive (for example Recital 40 and Article 7(1)), it seems that what the new right is aimed at protecting is the investment in the database's creation, and that investment must be directed at obtaining, verifying or presenting the database's contents.

However, the effort put into creating the data itself is irrelevant - Article 7(4) draws a distinction between rights in the database and rights in the data in the database.  Thus, the cost of BHB in fixing the date of a race meeting is not relevant to the question of investment, but the efforts that went into the gathering of all the data together were relevant.  Recital 55 effectively provides that, if sufficient investment is put into ensuring that it is up to date and accurate, it is protected by the new right.

Database right also protects the investment in presentation of data, which must cover at least the effort and resources that make the data more readily accessible by the user.  However, Article 7(3) provides that database right does not cover computer programs used to make or operate databases accessible by electronic means.  This may give rise to difficult questions as to where the exact borderline lies, but this question did not arise in this case.

The Claimant had to prove under Articles 7(1) and (5) of the Directive that the Defendant was misusing (in a relevant manner) parts of its database rather than data obtained from some other, independent source.  William Hill has on its websites information admittedly derived from the SIS RDF.  This is identical to information supplied to SIS by BHB from the BHB Database and also to the information supplied through the Racing Pages in the Declarations Feed which William Hill uses for its telephone betting service.  This was important, since William Hill argued that some at least of the information it took from the RDF could well have been obtained directly by SIS from racecourses and BHB had not proved otherwise.  William Hill pointed to the existence of an agreement between the Racecourse Association Limited and SIS, allowing SIS to broadcast races and obliging members of the Racecourse Association to provide SIS with race information in advance.

This argument was rejected on the facts.  The evidence showed that the racing world looked on the information derived from the BHB Database as definitive, and there was no reason why SIS would look to the less complete information delivered by individual racecourses.

Article 7(1) provides that one of the purposes of database right is "to prevent extraction and/or re-utilisation of the whole or of a substantial part, evaluated qualitatively and/or quantitatively, of the contents of the database".  BHB argued as follows:

  • what William Hill took by the RDF was a substantial part of the contents of the BHB Database;
  • although the BHB Database contains more information than what was actually taken, what was taken was core data;
  • it was BHB's investment in keeping that information up to date and accurate which William Hill was using for the purpose of its new commercial venture.

William Hill argued the following points:

  • what William Hill used was not a part (in the relevant sense) of the BHB Database;
  • even if it was a part, it was not a substantial part;
  • the use made of it did not amount to an "extraction" from the BHB Database;
  • it was not a re-utilisation of the BHB Database.

William Hill's essential argument was that there was a difference between a database and the information contained in it.  Simply to take the facts contained in a database could not be infringement of database right.  Any acts, to be infringing, would have to take advantage in some way of the arrangement of the contents of the database, or the way the maker has made them accessible.  William Hill further argued that database right, in effect, had to be construed as being narrower that the protection that used to be afforded to compilations under English copyright law.

This approach was incorrect.  While a collection of data had to have the form of a database as defined by the Directive, this did not mean that that was the limit of the protection.  It was not possible to take the contents of a database, re-arrange them and thereby avoid infringement of the new right.

As to substantiality, William Hill argued that, of the total data contained in the BHB Database, very little, relatively speaking, had been taken.  This approach was wrong.  While the test was not simply to gauge the importance to the Defendant of what had been taken, that might throw some light on whether something substantial had been taken.  Article 7(1) requires one to look at both qualitative and quantitative criteria, but that did not mean looking at them separately, they had to be looked at in combination.  In this case, the Defendant was making use of the most recent and core information in the BHB Database, and thereby relying on the completeness and accuracy of that information taken from the RDF i.e. the product of BHB's investment in obtaining and verifying that data.  This was a substantial part of the contents, and it made no sense to look at the issue first on a quantitative basis, and then on a qualitative basis.

"Extraction" and "re-utilisation" are defined very precisely in Article 7(2):

"(a) "extraction" shall mean the permanent or temporary transfer of all or a substantial part of the contents of a database to another medium by any means or in any form;

(b) "re-utilisation" shall mean 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.  The first sale of a copy of a database within the Community by the right holder or with his consent shall exhaust the right to control resale of that copy within the Community; public lending is not an act of extraction or re-utilisation."

William Hill argued for a restrictive interpretation of each expression.  As to extraction, William Hill argued that it was only the first removal of the data, after which infringement could only be committed by a re-utilisation of that data.  This argument was rejected.  "Re-utilisation", according to William Hill, meant making available to the public for the first time.  This argument was also rejected.

In conclusion, William Hill's action of taking information from the RDF and loading it onto its own computers for the purpose of making it available on its website was an unlicensed act of extracting a substantial part of the BHB Database and the subsequent transmission of that data onto its website for access by members of the public was a re-utilisation.

BHB also relied on Article 7(5) which states:

"The repeated and systematic extraction and/or re-utilisation of insubstantial parts of the contents of the database implying acts which conflict with a normal exploitation of that database or which unreasonably prejudice the legitimate interests of the maker of the database shall not be permitted."

William Hill argued that what it was doing was taking very small bits of information from what was a constantly changing database.  This amounted to taking one insubstantial part from a sequence of different databases, each of which was a new database coming into existence as ever changing details of horses, races and so on were added.  While this was an intriguing argument, it was also rejected.  It was impossible and unrealistic to try to split the BHB Database into separate databases.  As new data was added, the database's term of protection was constantly being renewed.

William Hill also argued from the closing words of Article 7(5) that its activities did not conflict with normal exploitation of the BHB Database and did not unreasonably prejudice the legitimate interests of BHB.  What amounted to normal exploitation or unreasonable prejudice was perhaps impossible to define, however, William Hill's activities clearly undermined a significant part of BHB's exploitation of its database and unreasonably prejudiced its interests, since it risked significantly affecting the value of the licences for the use of the database.

Finally, William Hill had proposed to circumvent any negative ruling by the Court by making some amendments to its website.  If it could not publish the race names or times, it could give the number of the race instead.  However, it was held that this was simply manipulating information that had already been wrongfully extracted.  Such manipulation itself was infringing by re-utilisation.


Many people were confused (and remain confused) by the introduction of database right.  This is the first case (and the only case in this jurisdiction at the time of writing) looking at this new right, and it is therefore valuable in giving some guidance as to the scope of the right.  In that sense, it  is useful that the judge took the course of outlining the main features of the right.  The most important principle to emerge from the judgment is that database right, whatever it is, should be interpreted by reference to the principles underlying the reasons for its creation, and it is not necessary to import from copyright law principles and features which are specific to copyright law.

Of interest to those who are otherwise unfamiliar with modern intellectual property decisions, is the judge's practice of quoting only from the original Directive, and not from the implementing legislation (in fact the Copyright and Rights in Databases Regulations SI 1997/3032).  The latter is in fact normal when judges of the Patents Court look at rights deriving from European legislation.

With the full report of the judgment given above, it barely seems necessary to look at the actual result: indeed, the arguments put forward by William Hill seemed to have more of ingenuity to them than practicality and so the result on that basis seems to be well founded.  It is perhaps worth setting out the basic core of this new right with an eye on the judge's words in this case.

It is probably true that UK copyright law was not in need of intervention in this area, and that the real thrust of the Directive was aimed at other European countries whose laws did not provide for effective protection of databases.  UK copyright law, which had provided protection for compilations, was less obviously in need of refinement in this area.  Unfortunately, the result of the Directive is something of an ugly mess, akin to the clash between designs and copyright.  The reason for this is that there is no clear dividing line between copyright and database right: it is possible for one and the same database to benefit from both rights.  However, the scope of protection given by each right is different.

A database is defined by Article 1(2) of the Directive as "a collection of independent works, data or other materials arranged in a systematic or methodical way and individually accessible by electronic or other means."  Foreign corporations should note the qualification requirements in Paragraph 18 of the Regulations, basically limiting database right to EEA nationals and corporations.  However, the principal difference between copyright and database right emerging is that the only criterion for copyright protection is that the database is the author's own intellectual creation.  For database right, Article 7(1) of the Directive provides that the new right applies to a maker of a database who "shows that there has been qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents".  The case of William Hill of course looks at this concept, and the judge is careful to set out in some detail the complexity and refinement of the BHB Database as well as the considerable work that goes into its verification.

One difficulty with databases is that they are rarely static, and this impacts on their duration, which is normally 15 years from the end of the calendar year in which the making of the database was completed or when it was first made available to the public.  This is somewhat troublesome.  There is a sense in which most databases are never finished.  For example, if I start a comprehensive catalogue of commercially available CDs, this is a task that will never finish (assuming manufacturers continue with the CD format), so when does the 15 years start from?  Perhaps only when I have catalogued albums by ZZ Top. 

Taking that example forward, if I keep the database up to date with the latest releases, what happens to the term then?  This is an issue that caused problems during the drafting of the Directive and a solution was attempted in Article 10(3):

"Any substantial change, evaluated qualitatively or quantitatively, to the contents of a database, including any substantial change resulting from the accumulation of successive additions, deletions or alterations, which would result in the database being considered to be a substantial new investment, evaluated qualitatively or quantitatively, shall qualify the database resulting from that investment for its own term of protection."

This is fine as a statement of principle, but its application in practice is fraught with difficulty.  There was little problem in William Hill, since the continuing work was so intense, that changes could be said to be substantial.  Going back to the example of the CD catalogue, what level of work would be required to make substantial changes?  What if I go through bands and composers on a purely alphabetical basis over the course of years, when does the right vest, and when would it be renewed?  Such questions remain unanswered (and possibly unanswerable), although Recitals 53 and 54 of the Directive are relevant here, as they put the burden of proof on these issues on the maker.

The infringing acts are also different from copyright, consisting of "extraction" and "re-utilisation".  These terms are defined in the Directive and are discussed in the judgment.  Note that the judge in William Hill shied away from applying a two-stage test of quantitative and qualitative elements.  The judge's comments on Article 10(3) also deserve note, in particular his rejection of the somewhat scholastic argument that the BHB Database was changing so frequently that repeated extractions could not be from the same database.  This argument is reminiscent of Herclitus' statement that one could not step into the same river twice.

The Directive tries to get around the problem of where an individual act that potentially infringes is only of an insubstantial amount, but it happens repeatedly over a course of time.  Article 7(5) also received comment from the judge, as it provides:

"The repeated and systematic extraction and/or re-utilisation of insubstantial parts of the database implying acts which conflict with a normal exploitation of that database or which unreasonably prejudice the legitimate interests of the maker of the database shall not be permitted."

This problem had arisen in copyright law.  In Cate v Devon & Exeter Constitutional Newspaper Co (1889) 40 Ch D 500 the defendant took small amounts from the plaintiff's newspaper at regular intervals and included it in his own newspaper.  The finding was that he had infringed copyright, even though the amount taken each week was, seen by itself, actually very small.  Curiously, Laddie J. cast doubt on this in Electronic Techniques (Anglia) Ltd v Critchley Components Ltd [1997] FSR 401, so the matter seems unsettled at least in copyright law.

The matter is important in database right, since infringement only happens when the infringer extracts or re-utilises all or a substantial part of the contents of a database - it is expressly not an infringement for a lawful user to extract or re-utilise an insubstantial part (Article 16(1) of the Regulations and Articles 7(2) and 8(1) of the Directive; see also Recital 49).

The result of this case is not great new law, but an interesting application of otherwise confusing principles.  Given the importance of databases for all manner of on-line activities, it is a case that, however, deserves careful notice.

We are processing your request. \n Thank you for your patience. An error occurred. This could be due to inactivity on the page - please try again.