England and Wales Patents County Court

IN THE PATENTS COUNTY COURT

		133-137 Fetter Lane London, EC4A 1HD
		20th June 2011

JUDGE BIRSS QC:

1. This action concerns database rights. The parties are rival producers of healthcare information. The claimant publishes a database called Binleys Database of GP Practices. It consists essentially of the names and addresses of individuals, such as practice nurses and doctors, associated with GP Practices. It has been described as providing "an accurate and reliable means to communicate sales and marketing messages to GP Practices throughout the UK" and the claimant has been publishing it since 1994.
2. Edition 29 of the database, referred to in the evidence of Mr. Brinzer, dating from about 2009, names 159,576 individuals located at 11,480 GP practices within the United Kingdom. Mr. Brinzer explains that the database is kept up to date and the claimant spends approximately £110,000 a year annually doing so.
3. The first defendant provides direct marketing information in the fields of list sourcing, print management, mailing and fulfilment, database management and campaign response handling. The second defendant carries on business providing database marketing services to manufacturing service companies in the healthcare sector.
4. The action is a claim for database right infringement under Regulations 12 and 16 of the Copyright and Rights in Databases Regulations 1997 SI 1997/3032 ("the Regulations") implementing Directive 96/9 ("the Directive"). Database right is sometimes called sui generis database right to distinguish it from the copyright that may also subsist under the Copyrights, Designs and Patents Act 1988.
5. In August 2007 the claimant found clear evidence that the first defendant was using information from the claimant's database. This was because the claimant puts a few seeds in its database. The seeds are dummy entries which do not correspond to real people. They are fictitious entries with addresses corresponding to the claimant's staff. The claimant will find out if someone is using data from its database because a letter will be sent to one of its seed addresses. That is what happened in this case and the action ensued.
6. The underlying facts were not in dispute. A mailing to the seed entry took place in about August 2007 when the defendants were conducting a mailing exercise of a book called the OTC Directory for GP Practices. The packets containing the book were addressed to individuals such as practice nurses and GPs in the GP Practices. The project was being conducted for a company called Communications International.
7. The packets were sent by the first defendant. The first defendant obtained the data used from the second defendant and the second defendant in turn obtained the data in March 2006 from an organisation called Bespoke Database Organisation Limited ("BDOL"), associated with a Phillip Bodwell. The data from BDOL included the seed entry. There is no doubt that the BDOL data includes at least one entry from the claimant's database, that is the seed. There is also no doubt that BDOL used the claimant's database as one of the sources from the BDOL data. As I shall explain in more detail below, the issue for trial is whether the data used by the defendants amounts to a substantial part of the claimant's database.

The Proceedings

8. The claim form was issued on 20^th February 2009. In October 2010 I heard an application for summary judgment, there were two points: First, had the action settled? The claimant said not. Second, if not, the claimant sought summary judgment on the whole claim. The outcome of that application is my judgment at [2010] EWPCC 12. I decided that the action had not settled. I gave summary judgment on title and subsistence of database right, but not on the question of infringement. Infringement was denied. The defendants pointed out that all that had been shown was the presence of a single seed entry in the defendants' data. They denied there was evidence of a substantial part of the claimant's database having been taken.
9. However probable it might have been that more data had found its way from the claimant's database to the defendants', it was not possible to decide the issue as a question of summary judgment. I gave the parties time for settlement discussions and ADR to take place and gave directions for trial, the trial to take one day.
10. The trial came on before me. Mr. Hill appeared for the claimants instructed by Birketts and Mr. Power appeared for the defendants instructed by Ashton Bond Gigg. The issues are: first, the question of infringement; and, second, the question of the liability of the two defendants. On the latter point Mr. Caldwell-Nichols is the managing director of both companies, but they are run as distinct business and I need to deal with that.

The Witnesses

11. Walter Brinzer and Andrew Farmer were the witnesses for the claimant. Mr. Brinzer is the joint managing director of the claimant. He joined in 1993. He had an intimate knowledge of the claimant's database products, including this one. He described the work that had gone on to produce and maintain the database over the years. Mr. Farmer was the assistant head of production of the claimant since 2007. He had analysed the information provided by the defendants in disclosure in order to show the extent to which the data had changed over time. In my judgment, both of them were good witnesses and gave their evidence fairly.
12. Matthew Caldwell-Nichols was the witness for the defendants. He is the managing director of both companies. He explained the position of the defendants, how they acquired the data which led to this case and how they had used it. Mr. Hill criticised his evidence in two respects: The first point concerned his view (that is Mr. Caldwell-Nichols' view) on what inference one should draw from the significance of a third party's seed being found in the data that he was using. The second point related to a document which was in the defendants' disclosure, which appeared to show the defendants' customer (Communications International) had obtained a quotation from Binleys (that is the claimant) for in effect the same job as the one which gave rise to this case, albeit a few months after it had taken place. The claimant was suspicious that this was not a genuine business exercise but had been organised by the defendants in order to gauge their potential damages liability to the claimants once they had realised what had happened. The claimant also pointed out that the quantities in the quotation (33,640 GPs and 10,367 practice nurses) were different from the number of records the defendants had contended were used for the OTC Directory project: 6,000 nurses and 36,000 GPs making a total of 42,000 entries.
13. I reject both points as criticisms of Mr. Caldwell-Nichols. First, the issue, in one's opinion, about the significance of a seed is, in my judgment, entirely debateable. As to the second point, the position developed significantly on this after trial. I received a communication on the issue from the defendants' solicitors on 14^th June and a letter in response from the claimant's solicitors on 15^th June. Prior to receiving the post-trial letters my view on the matter was that, although the document was somewhat mysterious and the circumstances were a little unclear, nothing Mr. Caldwell-Nichols said about it seemed to me to be open to criticism. He could not explain the document, but he gave clear evidence about his recollection of what Communications International had asked his company to do. In my judgment, Mr. Caldwell-Nichols had given his evidence fairly.
14. The post-trial letters included a letter written after the trial by Mr. O'Brien of Communication International and copies of emails from the defendants. It is now tolerably clear that Communications International did indeed obtain a quotation from the claimant for the same job and did so in about May or June 2007. Thus as a point against Mr. Caldwell-Nichols' position as a witness, the quotation issue falls away in any event. What is plain, however, is that documents which ought to have been disclosed by the defendants have not been, in particular, the emails provided with the post-trial letter. They plainly ought to have been disclosed. Mr. Caldwell-Nichols was asked about disclosure and said he disclosed everything his solicitors had told him to. I am in no position to resolve that debate. The fact is that disclosure ought to have been given.
15. The claimant also submitted that the disclosure raised further questions of disclosure and related to the joint tortfeasance plea. I will deal with those points in context.

The Facts

16. The claimant has been publishing the database in question since 1994 and has maintained it ever since. The database started at the time as a database of GP Fundholder practices. In 1997 its name was changed to reflect its content as a database of GP Practices in general. In 1994 the first seed entry was created, a phantom GP practice. This consisted of a GP practice with an address using Mr. Brinzer's then home address. The entry includes the names of various phantom doctors, a practice nurse, a practice manager, a person responsible for the computer system and so on. In 1997 the address of the first seed was changed to a different address because Mr. Brinzer had moved house. In 1996 another seed entry was inserted. It is another phantom GP practice. This time using the address of the business manager of the claimant's mailing services department. Again, it includes phantom doctors, practice nurse, health visitor, practice manager and so on. That second seed was present in the database until 2008. I should note that one seed does not equate to a single record in the database. From the point of view of records of individuals, a single seed may include five or six individuals. It was the mailing of the OTC Directory to the practice nurse at the relevant address for the second seed which alerted the claimant to the problem in this case.
17. Following disclosure can be seen that inside the database used by the defendants there was both the first seed and the second seed. Furthermore, the first seed entry uses the earlier pre-1997 address. The only edition of the claimant's database with that combination is edition 5 of the database, which was current between February and Autumn/Winter 1997.
18. Now we know that the database used by the defendants came from Phillip Bothwell's company, BDOL. Mr. Brinzer explained that Mr. Bothwell had been in contact with the claimant since February 1997. In April 1997 he bought the practice nurse data from edition 5 of the Binleys database. The invoice for that purchase, reference 2654, shows that the data consisted of 8,363 records of practice nurses. A fee of £2,382 was paid for a licence for a single usage of that data. In other words, the data was sold in order for Mr. Bothwell to use it for mailing once. Mr. Bothwell was supposed to have deleted those records after use. He did not do so. He kept them. We also know that Mr. Bothwell has, over the years, bought other data from the claimant's database. By 2006 we know that Mr. Bothwell had himself a GP practices database consisting of some 180,000 records. It is clear that Mr. Bothwell's database had 8,000 odd practice nurse records in it. It may well be that there is more in that database from the claimant (I will return to that at the end).
19. From Mr. Farmer's evidence it is clear that 6,901 of the practice nurse records in Mr. Bothwell's database as at 2006 are an exact match to the claimant's edition 5 database. The most likely explanation is, and I so find, that all the 8,363 practice nurse records obtained by Mr. Bothwell from the claimant were put into his collection in about 1997 in breach of his licence. The reason that not all 8,363 records are identical to edition 5 today is because some have changed since. The data may have been processed with names being updated or whatever. I find as a fact that the 6,901 records in the BDOL database in 2006 were derived from the claimant's edition 5. Of the 14,000 records which are not identical it is probable that the majority of them are records derived from the claimant's edition 5 database, but with modifications.
20. In 2006 the second defendant bought BDOL's whole GP practices database for some £113,000 plus VAT. It is clear, and I find as a fact, that in 2006 the second defendant loaded the database they had obtained from BDOL on to the second defendants' computers. In doing so they loaded the practice nurse records on to their computers. It is also clear, and I find, that all of the edition 5 practice records in the BDOL database were loaded on to the second defendants' computers. So at least 6,900 edition 5 identical records, and probably all 8,363 records with modifications, were loaded on.
21. Then, in 2007, Mr. Caldwell-Nichols was asked to conduct the OTC Directory mailing exercise of Communications International. Mr. Caldwell-Nichols said that it was to send out to 42,000 GPs and 6,000 practice nurses. Mr. Hill did not accept that those numbers were right. He relied on the mysterious quotation obtained from the claimant that appeared in the defendants' disclosure to support a case that the total may be a little higher. It is not possible to resolve that debate and I will work on the assumption at this stage that the number of practice nurses was at least 6,000.
22. As I understand it, the commercial arrangements for this job were that the second defendant, Precision Direct Marketing ("PDM"), had the contract with the client (Communications International). PDM is the "customer facing entity" in Mr. Caldwell-Nichols' words. The first defendant, Guardian, physically produced the wrapped books. PDM provided the data to be used to Guardian. Guardian used it to print the envelopes to post out the OTC Directory books and this all happened in August 2007. So, by this process, Guardian loaded the data on to their computers to carry out the project.
23. The defendants have produced the 6,000 records that they say were used for the mailing. Based on Mr. Farmer's analysis, which I accept, 4,783 of the records from the 6,000 (that is 57%) are identical to records in the edition 5 of the claimant's database. The explanation for the difference is, again, likely to be the result of updating and processing work on the data. So I find that at least 4,783 edition 5 identical records, and probably all 6,000 records with modifications, were loaded on to Guardian's computers and were used to conduct the mailing. One of the letters was to the practice nurse at the second seed. It came to the claimant's attention and this dispute arose.

The Law

24. The provisions relating to infringement of database rights are set out in Regulations 12 and 16 in the Regulations. The relevant parts of those Regulations are:

12.—(1) In this Part

'extraction', in relation to any contents of a database, means the permanent or temporary transfer of those contents to another medium by any means or in any form;

'insubstantial', in relation to part of the contents of a database, shall be construed subject to Regulation 16(2);

're-utilisation', in relation to any contents of a database, means making those contents available to the public by any means;

'substantial', in relation to any investment, extraction or re-utilisation, means substantial in terms of quantity or quality or a combination of both."

Regulation 16:

(1) "Subject to the provisions of this Part, a person infringes database right in a database if, without the consent of the owner of the right, he extracts or re-utilises all or a substantial part of the contents of the database.

(2) For the purposes of this Part, the repeated and systematic extraction or re-utilisation of insubstantial parts of the contents of a database may amount to the extraction or re-utilisation of a substantial part of those contents."

25. In Directmedia GmbH v Albert-Ludwig Universitat Freiburg, Case C-304/07) the Court of Justice considered the scope of the concept of "extraction" in the context of database right. Article 7(2) of the Directive defines "extraction" and the relevant parts of Regs. 12 and 16 together are clearly intended to put that into effect.
26. Paragraph 34 of Judgment in Directmedia states that:

"the concept of extraction, within the meaning of Article 7 of Directive 96/9, must be understood as referring to any unauthorised act of appropriation of the whole or a part of the contents of a database (see The British Horseracing Board and Others, paragraphs 51 and 67)."

Moreover paragraph 37 of Directmedia decides that, for the purpose of assessing whether there has been an extraction, it is immaterial whether the transfer was based on a technical process or was a simple manual process. The Court therefore held that the right is not limited to preventing mechanical reproduction without adaptation of the database or a part of it (see Directmedia paragraph 40). The question is whether an extraction is in respect of a substantial part is evaluated qualitatively or quantitatively. The extraction may be in one go or by repeated and systematic extractions of insubstantial parts which in overall amount to a substantial part (Directmedia paragraph 59). So, if what was extracted was substantial on a quantitative basis, that would be infringement. Alternatively, even if what was extracted was not substantial on a quantitative basis but was substantial on a qualitative basis, that would also be infringement.

27. How does one assess what is substantial? The answer is provided by The British Horseracing Board v. William Hill C-203/02 [2005] RPC 13 (the BHB case) and, in particular, the passages from and including paras.68-82 inclusive. In particular, it bears quoting paras.70 on quantitative and para.71 on qualitative substantial parts as follows:

"(70) The expression substantial part, evaluated quantitatively, of the contents of a database within the meaning of Article 7(1) of the directive refers to the volume of data extracted from the database and/or re-utilised, and must be assessed in relation to the volume of the contents of the whole of that database. If a user extracts and/or re-utilises a quantitatively significant part of the contents of a database whose creation required the deployment of substantial resources, the investment in the extracted or re-utilised part is, proportionately, equally substantial.

(71) The expression substantial part, evaluated qualitatively, of the contents of a database refers to the scale of the investment in the obtaining, verification or presentation of the contents of the subject of the act of extraction and/or re-utilisation, regardless of whether that subject represents a quantitatively substantial part of the general contents of the protected database. A quantitatively negligible part of the contents of a database may in fact represent, in terms of obtaining, verification or presentation, significant human, technical or financial investment."

28. Thus, on the quantitative approach, I should consider the volume of data extracted as against the volume of data in the database as a whole. If a user takes a quantitatively significant part, whose creation involved substantial resources, then there is infringement. On the qualitative approach, considering the scale of the investment, I should consider the scale of the investment in obtaining, verifying or presenting the contents regardless of whether that amounts to a quantitatively substantial part in the whole. I should also note that the intrinsic value of the data is not relevant (see the Court of Justice's judgment, para.78).
29. Also in the BHB case, the court considered Article 7(5) of the directive, that is Regulation 16(2) of the UK Regulations. This is set out in paras.83-96 of the judgment, and the key passage is paras.88-91 as follows:

"(88) Similarly, Article 7(5) of the directive prohibits third parties from circumventing the prohibition on re-utilisation laid down by Article 7(1) of the directive by making insubstantial parts of the contents of the database available to the public in a systematic and repeated manner.

(89) Under those circumstances, acts which conflict with a normal exploitation of [a] database or which unreasonably prejudice the legitimate interests of the maker of the database' refer to unauthorised actions for the purpose of reconstituting, through the cumulative effect of acts of extraction, the whole or a substantial part of the contents of a database protected by the sui generis right and/or of making available to the public, through the cumulative effect of acts of re-utilisation, the whole or a substantial part of the contents of such a database, which thus seriously prejudice the investment made by the maker of the database.

(90) In the case in the main proceedings, it is clear, in the light of the information given in the order for reference, that the acts of extraction and re-utilisation carried out by William Hill concern insubstantial parts of the BHB database, as stated in paragraphs 74 to 80 of this judgment. According to the order for reference, they are carried out on the occasion of each race held. They are thus of a repeated and systematic nature.

(91) However, such acts are not intended to circumvent the prohibition laid down in Article 7(1) of the directive. There is no possibility that, through the cumulative effect of its acts, William Hill might reconstitute and make available to the public the whole or a substantial part of the contents of the BHB database and thereby seriously prejudice the investment made by BHB in the creation of that database."


30. The point of this part of the Directive is to prevent third parties circumventing the main prohibition laid down in Article 7(1), i.e. Regulation 16(1) So, in the BHB case, although William Hill were making systematic and repeated small extractions, they were not such as to circumvent Article 7(1) because there was no risk the database as a whole or a substantial part of it could be reconstituted. So those acts were found not to violate Article 7(5).
31. The parties also cited the Apis Hristovich v Lakorda case C-545/07 [2009] Bus LR 1554. The defendants pointed out that at para.74 of the judgment of the Court of Justice there was a confirmation that, in evaluating whether a quantitatively substantial part has been extracted, the comparison is between materials allegedly extracted and the total database from which it was extracted.
32. I should also record that, at one stage, Mr. Power appeared to me to seek to be relying on the fact that his clients had done all the acts complained of unwittingly. I do not need to resolve that question because it is clear, in my judgment, that the question of infringement of database rights is a tort of strict liability. The state of mind of the defendants is irrelevant to the question of liability.

Issues

33. I understand it to be accepted (and, in any case, I find) that the loading on to PDM's computer of the practice nurse records from BDOL was an act of extraction within the meaning of the database directive and the Regulation. I also understand it to be accepted, and in any case I find, that the loading on to Guardian's computer of the practice nurse records from PDM was also an act of extraction within the meanting of the database directive. In both cases the issue is, and is only, whether what was extracted was a substantial part (as that term is defined) of the database within which the claimant claims rights. I will turn to that below.
34. Also in issue is the question is whether the mailing process, by printing individual letters, amounts to unlawful extraction under Regulation 16(2) (or Directive 7(5)) as a repeated and systematic extraction of insubstantial parts which may amount to extraction of a substantial part. Mr. Hill made it clear that he did not maintain the argument that this amounted to re-utilisation, and I do not consider that.
35. Edition 5 of the database contained 43,000 records. Mr. Power submitted that the right comparison from the point of view of the mailing exercise was between the 6,000 practice nurse records stored on Guardian's computer and used in the exercise and the 43,000 records in edition 5. He also submitted that the right comparison from the point of view of the data held by PDM to have been bought from BDOL was between 8,000 practice nurse records stored on PDM's computer and the 43,000 records in edition 5. Mr. Power submitted that in each case the 6,000 or 8,000 records were not a substantial part, quantitatively or qualitatively. He submitted that I have no evidence to go on at all to address the crucial question of investment, which is the key to it. He submitted his best case as to the 6,000 figure since it is a lower figure. I will work on that assumption, although I note that Mr. Hill did not accept the 6,000 figure is correct.
36. In my judgment, Mr. Power was right to focus on the total number of records likely to be used, whether identical or somewhat modified, rather than only on the identical records. But I will also check whether the number of identical records, which is always lower, makes any difference.
37. I will turn to the question of investment. Mr. Power submitted that there was no proper evidence of investment so as to carry out the exercise. The evidence was from Mr. Brinzer. He had explained the general process by which the claimant's database was created, updated and maintained. He explained what that cost and the work that was involved. The figures Mr. Brinzer used were the position today. His evidence was (and I am quoting from para.5 of his witness statement) as follows:

"Our methodology for obtaining the database is as follows:

(a) First, we send a questionnaire by mail to approximately 13,000 establishments. This currently costs about £12,000 in mailing costs per annum. An example questionnaire is enclosed.

(b) Second, we use telephone research and contact each establishment on average four times per year or twice per cycle. Responses to these questionnaires are never complete and frequently contain inaccuracies, which means that we need to validate a certain amount of information each cycle by telephone. We consequently make about 52,000 phone calls per year. These calls enable us to get a lot of data that people do not volunteer in response to the mailed questionnaires, enabling us to build up a complete and comprehensive data set. Our reputation and competitive edge is built in completeness and quality of our data set. Telephone research is currently budgeted at about 18 man weeks for both cycles. We define a man week as 35 hours and the cost, including overheads, is £25 per hour. This comes to a total of £70,000 per year on telephone research.

(c) Third, we budget 32 man weeks for data capture, also at £25 per hour. This gives a total of £28,000 per year on data capture.

(d) Fourth, we commit another 2 man weeks to data quality checking. This work costs slightly more at £32 per hour. This comes to about £2,200."

Mr. Power put to Mr. Brinzer that there was no evidence of the position relevant to edition 5. Mr. Brinzer's position was that the figures were scalable with a number of records in the database at a given time, subject to inflation for financial figures. Thus, if one wanted numbers for edition 5, a simple pro-rata calculation would give them. I accept Mr. Brinzer's evidence. Edition 5 was 43,000 records, whereas, at the time of Mr. Brinzer's evidence, the database was about 150,000 odd records. Thus edition 5 is about 30% of the size of the database today.

38. Accordingly, considering the number of phone calls made per year, the figures are as follows: Instead of 52,000 phone calls per year, pro-rata about 15000 phone calls per year will have been made at the relevant time. Thus 8,000 practice nurses in edition 5 will represent 8/43rds of the total, which is just under 3,000 phone calls per year. 6,000 records would represent about 2,200 phone calls per year. The lowest number of all, 4,783 identical records present in the mailing list, would correspond to about 1,800 phone calls. One can do the same with the figures for the man weeks' worth of telephone research. 80 man weeks today (in Mr. Brinzer's evidence) pro-rata would represent about 24 man weeks at the time of edition 5 and 6,000 practice nurses from edition 5 would represent 3 man weeks per year for telephone research. I am only going to do this with the two numbers I have dealt with, since they illustrate Mr. Brinzer's point.
39. In my judgment, this shows that I do have evidence in relation to the relevant investment in terms of obtaining, verifying and presenting the contents of the database.
40. A key point made by the defendants related to the degradation of the data. They say that one must assess the value of the data at the time it is used, i.e. when it was extracted. There is no dispute that it was used in 2006/2007 and we can tell from Mr. Farmer's evidence that most of the records derived from edition 5 were, by then, inaccurate. This arises as follows: Mr. Farmer had explained that edition 5 had 8,363 practice nurse records dating from 1997. By 1999, edition 8 of the database, there were only 5,769 practice records in the database which were identical to the practice records in edition 5, a drop of 29% in two years. The point was that the claimant's processes of updating and verifying their data tracked real life. People change jobs and move, etc. This showed that the quality of the data, if no continuous updating process was employed, degrades over time. The defendants performed a linear extrapolation ten years on, to the point when the mailing exercise was conducted. By then only 20% of the records in edition 5 in the database as it was from 1997, would have been accurate as at 2007. So in effect, say the defendants, the data was valueless.
41. In cross-examination Mr. Farmer accepted this data would be valueless from the point of view of using the practice nurse names. However, the qualification to "names" is important. Mr. Farmer's analysis was performed testing the identity of the whole records, so, if the name of a practice nurse changes, it is no longer identical. But the address of a practice may very well not have changed. So the address data in the records still has a substantial value, even though the nurses' names have changed. For example, one could simply write to "The Practice Nurse" at the relevant address. Nevertheless, Mr. Power submitted that this degradation shows that the data actually extracted by the defendants when loaded from BDOL in 2006 and when used in 2007 for the mail shot was in effect valueless, and therefore, he submitted, there could be no infringement.
42. In my judgment, the degradation point is irrelevant. It is nothing to do with the substantiality test. The test is what was taken from edition 5. The fact that the commercial value of what was taken may be low when it was taken ten years later, cannot be relevant. The point may or may not be relevant to quantum, but, in my judgment, it cannot be relevant to substantiality. The database rights last for 15 years. It seems to me that if the part extracted would have represented a substantial part in 1997, when it was current data, it would still be a substantial part of that 1997 database 10 years later. It does not stop being substantial just because time has passed and the world has moved on. I also cannot help but note that it is all very well for the defendants saying the data was valueless, but they used that very data and I presume charged their client for its use. Although it is unnecessary for me to do so since I have held the point is irrelevant, I will state that in my judgment the dataset used by the defendants (which was derived from edition 5) was not valueless in 2007. I accept Mr. Farmer's point that the dataset, and in particular the addresses, still had substantial value even without using the individual names of the practice nurses.
43. In my judgment, the 6,000 practice nurse records extracted for use in the mailing exercise represent a quantitatively substantial part of the edition 5 database. My reasons are as follows.
44. In terms of the quantitative question, we are comparing 6,000 records out of 43,000 records, i.e. about 14%. Even if one considers the 4,783 identical records, the fraction is about 11%. I note that Advocate General Stix-Hackl in the British Horseracing Board case raised the question of whether a quantitatively significant part was to be assessed in relative or absolute terms (see para.74 of her opinion and the passage from paras.72-82). It is clear that the Court of Justice came down in favour of a relative test because para.70 of the judgment of the court refers to an assessment in relation to the volume of the context of the database as a whole. In my judgment nevertheless, "quantitatively substantial" cannot require the numerical majority of the records. It must refer to a part having substance. In terms of volume, 11 or 14% of the data must be at the lower end of what could be regarded as quantitatively substantial. It seems to me, however, that even 11% represents a significant part of the volume of the database whose creation required substantial resources. Those are the words in para.70 of the ECJ's judgment. At its absolute lowest, the identical records represent 1,800 phone calls per year. So from 1994 to 1997 that amounts to some 5,000 phone calls to obtain, verify and present this data. That is a substantial amount of work, in my judgment. It represents an economic investment, which, it seems to me, the database directive sets out to protect.
45. Alternatively, regardless of my conclusion from a quantitative point of view, I find that the 6,000 records or the 4,783 identical records are qualitatively a substantial part of the edition 5 database. It seems to me that the scale of the investment that these records represent, (as I have found, thousands of phone calls and staff working for weeks to compile and validate the data) is a significant human and financial investment. On that basis, regardless of whether the volume is or is not quantitatively substantial, it is qualitatively a substantial part of edition 5.
46. That is sufficient to dispose of the point. If I am right about 4,783 and 6,000 then the conclusion obviously also applies to the 8,000 practice nurses in the database from BDOL. In my judgment the act of extraction by the second defendant loading the BDOL database on to its computers was an extraction of a qualitatively and/or quantitatively substantial part of the database and was an infringing act. Also the act of extraction by the first defendant loading the extract of the data to be used in the mailing exercise on to its computers was an extraction of a qualitatively and/or quantitatively substantial part of the claimant's database and was an infringing act.
47. Although I have decided the point, I will now turn to two other matters which I have not taken into account in reaching my conclusion since they appear to be potentially controversial but which, in my judgment, ought to be relevant and would support the conclusion which I have reached. First, Mr. Bothwell. Mr. Bothwell was prepared to pay about £2,300 for a single use of the 8,000 odd practice nurses in 1997. Is this relevant? BHB makes it clear that the intrinsic value of the data is irrelevant. However, in my judgment, this data has little intrinsic value in the sense that the individual record can come from the public domain. It is not a secret who is the practice nurse of a local GP surgery. One can find out. What Mr. Bothwell was paying is indicative of the investment in obtaining, verifying and presenting the data itself.
48. Now, of course, that price cannot tell you how the 8,000 he bought compares numerically to 43,000 in edition 5, but, in my judgment, it does indicate that the 8,000 (and thus the 6,000 too at the best case for the defendants) are the product of a substantial investment and therefore represent a qualitatively substantial part of the database. Moreover, I do not think this is inconsistent with my rejection of the degradation point for the simple reason that Mr. Bothwell was buying the data when it was current in 1997. In my judgment, this point would support the conclusion, albeit that I have not relied on it in reaching my conclusion.
49. The second point. I repeat that I have made this decision without dealing with this second factor, which, in my judgment, ought to be relevant. The point is that the records extracted are the practice nurse records. They are a well defined subset from the larger database. Thus, from the qualitative point of view, the totality of what was extracted represents a qualitatively distinct class of record in the database. The fact that they are practice nurses does not matter, but, in my judgment, it is highly relevant that they are all (or almost all) records of that type and that the remaining records in the database are all or mostly of a different type (or different types). Grouping the records into types or classes is a key part of the point of having a database in the first place and to ignore this aspect seems to me to be wrong.
50. Mr. Hill was concerned that BHB holds that the intrinsic value of records is not relevant and, therefore, was not prepared to submit that this factor per se was relevant. He did submit it could have a bearing on the issue of investment. While I understand Mr. Hill's caution, given the Court of Justice in BHB, nevertheless, it seems to me that recognising that records can be of different types and can be grouped in different groups is not giving them an intrinsic value and is not prohibited by BHB.
51. I doubt the practice nurse records are any more or less costly to produce than the other records, but they are a distinct set. Grouping the data into distinct sets is part of the essence of the work which goes into creating these databases in the first place. In my judgment, it would be wrong to ignore the fact that what was taken was in effect all the records of the given type. Part of the work undertaken is to mark these records accurately (as practice nurses, GPs or whatever), precisely in order to facilitate their specific extraction from the whole 43,000 records.
52. Finally, on infringement: was the mailing exercise itself an infringement under Regulation 16(2)? There were 16,000 letters with the name and address printed on them that have been extracted from the database. Each individual letter is of course an insubstantial part of the whole database. The individual extractions are repeated, one after the other, and systematic. They add up to a substantial part, on my findings already. However, I am struck by the Court of Justice in BHB making clear that the purpose of this provision is to prevent circumvention of the main prohibition on Article 7(1), Regulation 16(1). They held that William Hill's repeated and systematic extraction of insubstantial parts was not circumvention because the database could not be reconstituted from it. Therefore they did not infringe. Neither side addressed this circumvention point. It seems to me that one could reconstitute the 16,000 records from the 6,000 letters before they were posted and so, with some diffidence, I find that the exercise of writing the address labels is itself an infringement.

The liability of the two Defendants

53. The mailing exercise itself was done by the first defendant, in the sense that they committed the act of extraction. However, the second defendant gave the database to the first defendant for it to be used in the mailing. In my judgment, both defendants have infringed database right in the sense that each has loaded on to their own computers a substantial part of the relevant database. Also, in my judgment, the second defendant is liable for the mailing exercise since it extracted the mailing list to be given to the first defendant for use in the mailing. Thus both first and second defendants are, in their different ways, responsible for the mailing exercise.
54. However, Mr. Caldwell-Nichols denied that they were jointly liable for any tort committed by the other. Mr. Caldwell-Nichols stated in his witness statement that the two companies had not collaborated and Mr. Hill did not challenge that assertion in cross-examination. In my judgment, collaboration is not the issue. Based on the judgment in the Court of Appeal given by Peter Gibson LJ in SABAF v. Meneghetti [2002] EWCA 976, I take the law to be that the underlying concept is that the party must make the infringing act his own in order to become liable and the presence of the common design, concerted action or other combination to secure the act or make the combiners liable.
55. In this case it seems to me to be plain that the OTC mailing project of Communications International was a single common project and product of the concerted action of the two defendant companies. They were acting in combination to carry it out. In the paragraph of the first defendants' defence, which denies the plea of joint enterprise, the first defendant states that the mailing was organised and implemented by the first defendant alone. I am not satisfied that that statement is correct. In the emails produced after the trial by the defendants, which plainly should have been produced before, the claimant has identified a mailing budget document headed "Precision Direct Marketing", i.e. the second defendant, but contains a budget for the mailing exercise itself, i.e. the work said to have been organised and implemented by the first defendant alone.
56. The two companies plainly worked together to a significant extent on this project, but the details and extent of that work are wholly unclear. The budget document produced post-trial seems to me to bear out the claimant's concern about disclosure on this issue. In my judgment, the correct analysis is as follows. Although the legal burden of proof on this issue will be on the claimant, a prima facie case of joint tortfeasance exists. For one thing the budget document is prima facie evidence on the point. I am not satisfied that I have been given a complete picture of the relationship and dealings between the two companies in this project and that evidence is of course uniquely available to the defendants themselves. The evidential onus has shifted to the defendants on tortfeasance. It has not been discharged. I find the defendants are jointly and separately liable with each other as joint tortfeasance.
57. Finally, I need to deal with a very late disclosure application made by the claimant. The main thing sought is a full copy of the whole database purchased from BDOL in 2006. The point is that the claimant does not accept that the 6,000 odd practice nurses, the extraction of which is in effect admitted by the defendants, albeit not admitted to be a substantial part - is the totality of the claimant's database records which are present in the defendants' database. There was no application to adjourn the trial. It was plain that, even if I had ordered the disclosure first thing in the morning, the material could not have been produced during the one day trial. I decided not to de-rail the trial, but to hear the case and then decide what to do with the application. In my judgment, the practice nurse data infringed. It is also clear that the claimant does not accept that the practice nurses represent the full extent of the infringement. I have approached the issue on the defendants' best case, that the 6,000 practice nurses are the extent of the potential infringement. In my judgment, there is little doubt that more than just practice nurse data was passed to Mr. Bothwell over the years and, since the claimant was the source of part of the database sold in 2006, it is quite likely that more of the claimant's data was in the defendants' database when they created it by buying it from Mr. Bothwell in 2006. How much more may be relevant to relief.
58. Mr. Power told me at the conclusion of the case that it would take about 7 days to produce the material sought on the application. I will make an order now to produce the disclosure sought in 7 days from now. I will add that it seems to me that the claimant's point in their letter of 15^th June, that the disclosure of the totality of the mailing list used for the OTC project should be given, is a good one. I will hear the defendants on that point. That is my judgment.
________________