The assessment of quantum for copyright infringement in the context of IT disputes is a much maligned and often uncertain area. More often than not, the position is complicated by the fact that a piece of infringing code makes up just a small portion of the software in question. How is quantum to be assessed in those circumstances? What options are available to a claimant and how is a defendant to assess its potential exposure?
Infopaq International A/S v Danske Dagblades Forening [2009] EUECJ C-5/08, Public Relations Consultants Association Ltd v The Newspaper Licensing Agency Ltd and others [2013] UKSC 18 and other recent cases have shown that the threshold for infringing copyright can be surprisingly low. The test to be applied is qualitative, not quantitative. What may therefore appear to be a reproduction of only a small portion of someone else’s code can have serious consequences if the copied part contains elements which are the intellectual creation of the author of that code. The reality is that copying code is often the result of sheer laziness, but programmers who might have been inclined to take the easy option should proceed with caution – what might previously have been considered an ‘insubstantial’ piece of code, could be sufficient to constitute copyright infringement if it meets the Infopaq test. Calculating the appropriate compensation for a claimant and assessing potential exposure for a defendant in these circumstances can be a minefield!
In this article we look at how quantum is typically assessed after there has been a finding of copyright infringement and the potential impact for the parties in a software infringement dispute. It is important to bear in mind that the approach to assessing quantum in English law is underpinned by the equitable principle of unjust enrichment. The main question that must be considered is whether the defendant was unjustly enriched as a result of its use and exploitation of the infringing code. We will illustrate how the different assessment methods work in the context of an IT dispute by reference to our hypothetical case study.
We also consider how the assessment of quantum sits with the Directive on the Enforcement of Intellectual Property Rights (2004/48/EC) (the IP Enforcement Directive) and the potential impact this could have on defendants and claimants alike.
Basis of assessment
Like other registered and unregistered IP rights, it has long been established in English law that there are two ways to assess quantum for copyright infringement:
· inquiry as to damages; or
· account of profits.
These remedies are of course mutually exclusive: a claimant must elect for one or the other.
Damages
Almost 50 years ago, the House of Lords set out the principles applicable to assessing damages in the well-known patent infringement case of General Tire and Rubber Company v Firestone Tyre and Rubber Company Ltd [1976] RPC 197. In 2002, those principles were applied to copyright infringement disputes by the Court of Appeal in Blayney v Clogau St David’s Gold Mine [2002] EWCA Civ 1007.
In line with these authorities, the courts have adopted two main approaches to the assessment of damages:
A. The licence fee approach
Where there is evidence that the claimant has granted licences to others before or after the infringement, the courts’ approach is to objectively assess what royalty the infringer would have paid had it obtained a licence from the claimant. The court will ask what would a licence fee have been at the time of the infringement? Also, would it have been practicable for the infringer to license the copyright from the claimant at all, for example, if they were competitors? For a defendant in that position, it may be useful to ascertain how the claimant has behaved towards other competitors in the market and to get an understanding of its approach to licensing and use this as evidence to support a lower award, if possible.
If the claimant has not granted licences, the courts will be guided by rates charged in similar transactions in the relevant sector at the relevant time: a notional royalty rate under a notional licence.
It will also be important to work out how many licences the defendant would have needed to create the infringing software, and the total value of such licences would need to be estimated.
B. The lost sales plus user royalty approach
Where a claimant sells software incorporating the allegedly infringed copyright, the amount of damages will typically be based on the loss of ‘profits’ caused to the claimant by the infringement ie the sales that would have been achieved by the claimant but for the defendant’s infringing activities less any expenses that claimant would have incurred in creating the goods/providing the services (including a share of the general overheads of the business). This will involve consideration of a number of factors, including how competitive the market was at the relevant time, how much revenue each individual sale would have generated and what the value of any deductions would have been.
This approach to calculating damages may not be suitable where the infringing code comprises only a small element of the software in dispute and there are other elements unconnected to the claimant’s code, which have been used to create the defendant’s software product. Arguably it would be unfair to award all of the profit from the infringing sales to the claimant in those circumstances. Conversely, it could also be argued that the defendant would not have had such an attractive product to sell had it not made use of the claimant’s copyright.
In this situation, the amount of damages awarded should reflect the value of the infringing component only. Dissecting a finished product in this way is not an exact science. The court might also disagree with this approach as a matter of policy given that the claimant’s work has been copied and used in a finished product which was then commercialised to generate sales for the infringer.
Whatever approach is adopted, putting forward detailed evidence is crucial. It is clear from the case law that a failure to do so will be looked upon unfavourably by the court. For example, in Allen v Redshaw [2013] EWPCC B1, the judge remarked that neither party had submitted sufficient evidence as to sales figures or profits that the claimant might have made had the defendant not infringed the claimant’s copyright. Consequently, the judge was left with no option but to make assumptions which did not necessarily reflect the true position. There is also the possibility that failing to present proper evidence to the court at the appropriate time will result in a lower award for a claimant. Indeed, in Blayney, the court stated that it should err on the side of under-compensation where there is insufficient evidence rather than assessing damages liberally.
Claimants should also be wary of being seen to over-egg the pudding! Exaggeration of claims is frowned upon by the courts in any type of claim. One extreme example of this was in the case of Lilley v DMG Events Ltd [2014] EWHC 610 (IPEC), were the claimant put his damages claim for copyright infringement at a staggering £800 million. The court did not accept the claimant’s evidence/calculations of supposed licence fee charges as an appropriate guide to damages. The claimant’s figures had been calculated on the basis of royalty payments that he insisted the defendant would have paid but the figures did not correspond with the terms on which the claimant had previously granted licences. The court went on to strike out the claim after concluding that the maximum amount which could have been awarded was around £83. The court’s findings were based on an objective assessment of what would have been paid by way of a royalty had the claimant been a willing licensor and the defendant a willing licensee. Of course, that case is an example of serious overestimation and common sense prevailed.
The possibility of an award of additional damages under s 97(2) of the Copyright Designs and Patents Act 1988 (the CDPA) should also be borne in mind. The wording of the legislation shows that this is a discretionary remedy (‘The court may…’) which will be influenced by the flagrancy of the infringement and any benefit resulting to the defendant. As noted above, the aim of any award of damages is not to punish the defendant, but to put the claimant back in the position it would have been in but for the infringement. Therefore, whether a judge will exercise his discretion to award additional damages will depend on the circumstances of the particular case and whether the defendant would benefit/be unjustly enriched if an additional damages award was not made. In our view this would require evidence of particular bad behaviour on the part of a defendant which caused the claimant prejudice. There has been some recent judicial consideration of this (by His Honour Judge Hacon in Absolute Lofts Ltd v Artisan Home Improvements Ltd [2015] EWHC 2608 (IPEC)) and also as to whether s 97(2) of the CDPA has been made redundant by the IP Enforcement Directive (see below).
Account of Profits
Now we turn to the second basis of assessment. Like damages, the concept of an account of profits is underpinned by the equitable principle of unjust enrichment. The aim of an account of profits is to take the profits made by the defendant by reason of the infringement and transfer these to the claimant.
According to the court in Celanese International Corp v BP Chemicals [1999] RPC 203 it is incorrect to use an incremental approach when determining the amount of profit, ie simply to look at the difference between the profits made using the infringing code and the profits which would have been achieved by not using that code. Rather the correct approach is to look at the percentage of profits made by the defendant as against total revenue and deduct all relevant overhead costs. The level of profit must then be apportioned by reference to the infringing element of the code. In order to do so, typically it is necessary to obtain disclosure from the defendant.
In some circumstances it may be possible for the defendant to argue that it has not been unjustly enriched beyond the value of the licence fees for the infringing software that it would have had to pay to the claimant. This is because, had the defendant acted lawfully and bought licences from the claimant at the outset, the defendant could have gone on to create the same software anyway. In that event, the impact on the defendant’s profits would have been limited to the value of the licence fee.
Making an election: Case study
To put the above into context and illustrate the potential for differing outcomes between the different assessment methods, we will use a short case study: A creates software for use by medical practitioners in the diagnosis of illness. It has been doing so for five years and considers itself the market leader. B has been working in the same field for about two years. Six months ago, one of A’s lead developers left to join B. B has recently launched its own software product for use in the same field. According to A, part of B’s software product infringes the copyright in A’s software. B’s position is that the products simply function in a similar way, but there has been no copying. The parties are locked in a bitter dispute. They are weeks away from trial and attempts to resume settlement discussions are afoot.
A knows that there are approximately five competitor products to A and B’s software on the market. The market prices for a perpetual licence for those products tend to range from £300 to £800 per licence.
B’s software was used on 200 projects for 80 clients during the relevant time period. The total revenue from B’s 200 projects was £2,000,000, the average revenue being £10,000 per project.
Licence fee calculation
A’s licence fee for its product is £500 for a perpetual licence per user. B has 15 developers in its development team. Therefore, the maximum that B would have bought from A had it created its software product using A’s code (with the rights to do so) was 15 licences.
Calculating damages firstly by reference to the licence fee approach:
15 licences x £500 licence fee = £7,500
User royalty plus lost sales calculation
If, however, it is not accepted that A would have licensed its product to B (eg because they are competitors), we must consider which of B’s projects would have been awarded to A instead of to B. This is not easy to estimate and will depend, amongst other things, on where A ranked among the other five competitors in the market at the relevant time. Let’s say that A would have obtained approximately 1/2 of those 200 projects with approximately £10,000 revenue per project. This would put the total lost opportunity for A as roughly:
£10,000 revenue per project x 100 projects = £1,000,000
Next, we would need to assess the income attributable to the software after deduction of costs of sales and for provision of services. This will be a matter for evidence but let’s say in this case, the software income would be 30%.
30% of £1,000,000 = £300,000
A portion of general overheads should then be deducted from this. For the purposes of illustration, if the value of such overheads is a further 5% of those 100 projects and 2% of A’s other projects which are valued at £5,000,000. This would give the true profit level of such projects for A:
£300,000 – (5% of £1,000,000) – (2% of £5,000,000) = £150,000
Looking at the user royalty plus lost sales approach, if it is the case that the going rate for licensing A’s software was £7,500 for 15 licences, it is probably not the case that A would receive more than that on the user royalty basis because 15 licences could arguably have been used to create all 200 projects. Therefore, this £7,500 should be added to £150,000 to give £157,500.
Therefore, based on the ‘licence fee’ approach, the measure of damages for A would be £7,500. By contrast, the award on the basis of a ‘lost sale + user royalty’ approach would potentially be much higher at £157,500.
Account of profits calculation
An alternative approach is to look at how an account of profits would be assessed. In order to do this, it is necessary to consider what income, after deductions for costs and a portion of overheads, B made on the 200 projects. It will then be necessary to apportion an element of that value to the use of the infringing code.
If we assume that B had total revenue of £2,000,000 for all 200 projects and, after deductions for the cost of sales/provision of services and general overheads, the profits from the software component of those projects was 25%, this would give the following:
25% of £2,000,000 = £500,000
But if that profit cannot be entirely attributed to use of the infringing software (because it only comprises a small part of B’s product), an assessment of value of the infringing software (as part of the total software) must be made. One useful way of doing this might be to look at it in terms of working days and try to apportion the profit in that way to the infringing software. For example, if B’s product took 30 working days to write and 1.5 days of those days were spent on the code that is alleged to have infringed, one could argue that 5% of the actual profit made by B is attributable to the infringing code:
5% of £500,000 = £25,000
Although this is a somewhat simplistic example, it demonstrates how the parties might approach the assessment of quantum in a software infringement dispute. It also shows how difficult such calculations can be and therefore the importance of detailed evidence in respect of financials and sales to support any calculations.
The IP Enforcement Directive
Although consideration of injunctive relief is outside the scope of this article, it is worth mentioning the provisions of the IP Enforcement Directive, in particular Article 12. This states that the competent judicial authorities may order pecuniary compensation to be paid to the claimant instead of applying measures such as injunctions where the defendant acted unintentionally and without negligence, if execution of the measures would cause disproportionate harm and pecuniary compensation ‘appears reasonably satisfactory‘. The courts are increasingly questioning whether ordering an injunction in such circumstances would be proportionate, dissuasive and effective by reason of the IP Enforcement Directive.
Similarly, in the context of assessing damages, Article 13 of the IP Enforcement Directive provides that when judicial authorities set damages where the defendant has knowingly engaged (or had reasonable grounds to know it was engaging) in infringing activity, the defendant must pay damages ‘appropriate to the actual prejudice suffered….as a result of the infringement’. Such damages should take into account all appropriate aspects, including any negative economic consequences, lost profits of the claimant, unfair profits made by the defendant and, in certain cases, non-economic factors such as the moral prejudice caused to the claimant. This echoes the fact that assessment in this context in English law is based on the principles of unjust enrichment. As mentioned above, the co-existence of Article 13 with s 97(2) of the CDPA was recently considered by the Court in Absolute Lofts. Although not a dispute about software, in that case, HHJ Hacon held that Article 13 is wide enough to cover the question of awarding aggravated damages similar to those available under s 97(2) of the CDPA. He went on to award additional damages to the claimant in respect of the indirect benefits the defendant gained from its infringement, ie the increased business it received by using the claimant’s images on its website. The judge inferred that the claimant’s images made a contribution to the defendant’s profits between 2011 and 2013 by encouraging those who visited the defendant’s website to engage the defendant’s services and he noted that the contribution to that of the infringing material was ‘more than negligible’. The judge regarded the profits made by the defendant as particularly unfair because they resulted from a misrepresentation to the defendant’s customers that the work shown in the infringing images had been completed by the defendant when that was not the case. He was satisfied that the defendant’s infringement was flagrant. However, it is worth noting that, unlike s 97(2) of the CPDA, flagrancy is not a compulsory factor under Article 13.
In conclusion, HHJ Hacon confirmed that the amount awarded would have been the same in monetary terms whether it was calculated under Article 13 or s 97(2). The judge made it clear, however, that there was not an intention to sweep s 97(2) into the ambit of the IP Enforcement Directive. Rather, a successful claimant would be entitled to rely on either s 97(2) or Article 13, whichever would provide the greater damages.
Another point to consider is the potential impact of publication orders. Article 15 of the IP Enforcement Directive introduced a new remedy which may be granted in favour of a successful claimant as follows:
‘Member States shall ensure that, in legal proceedings instituted for infringement of an intellectual property right, the judicial authorities may order, at the request of the applicant and at the expense of the infringer, appropriate measures for the dissemination of the information concerning the decision, including displaying the decision and publishing it in full or in part. Member States may provide for other additional publicity measures which are appropriate to the particular circumstances, including prominent advertising.’
The rationale for making an order under Article 15 is to deter future infringements through the fear of further adverse publicity. The concern about such an order being made can be an extra incentive for a defendant to settle before trial.
Although there is no similar statutory provision in English law to grant such a remedy in the other direction, ie in favour of a defendant who has successfully defended an infringement claim, there is judicial discretion to extend the terms of an injunction granted under section 37(1) of the Senior Courts Act 1981. An example of a high profile case where this happened is Samsung Electronics (UK) Ltd v Apple Inc [2012] EWCA Civ 1339. Apple was required extensively to publish (online and in numerous UK newspapers) the decision against it that Samsung had not infringed Apple’s Community registered design. It was noted on appeal that one of the reasons such publication was necessary was to ‘dispel commercial uncertainty’ in the marketplace, either with the non-infringer’s customers or the public in general. However, the Court of Appeal stressed that any publication order made must be proportionate. Although, as noted above, assessment of proportionality can be subjective and is ultimately in the hands of the judiciary.
Formulating an offer
Formulating offers to settle in the context of copyright infringement cases can be challenging. The regime under Part 36 of the Civil Procedure Rules provides a mechanism for parties to try to settle disputes by placing pressure on the party in receipt of an offer to better that offer at trial. If a party rejects a Part 36 offer and goes on to get a worse result at trial, it will be penalised in costs. Normally, whether or not a party has ‘beaten’ an offer can be assessed purely in relation to the monetary value of the offer made and any damages award made at trial. However, copyright claims are more complex than that. In order to make such an offer and ensure it is compliant with the terms of Part 36, defendants need to explain the rationale for the offer. This may involve divulging potentially sensitive commercial information about profits to the claimant who may (most likely) be a competitor. An offer may also require a party to submit to an injunction. This invariably makes the settlement of copyright disputes challenging. However, on balance, settlement can be more attractive commercially than the potential ambiguity about the quantum of any court ordered damages award.
Conclusion
In the context of software disputes, there is a genuine concern that the threshold for copyright infringement has been lowered since Infopaq and subsequent cases. Arguably every aspect of written code is an expression of the author’s intellectual creation, and there is therefore a risk that reproducing what might previously have been considered an ‘insubstantial’ piece of code (previously not sufficient to constitute copyright infringement), might now be viewed differently by the English courts.
It is typically difficult to make an assessment about the value of an infringing component, in terms of its relative contribution to a body of code. If, for example, the issue relates to four lines of code within 40,000 lines of source code, it becomes a challenge for parties to assess that ‘contribution’. There may be only one way to write the code, or it might not have any value at all because those four lines might not do anything. Or it may be the single creative, elegant solution to the problem that the software was built to resolve. What is clear is that lawyers and/or judges cannot make a judgment about this by reading the code. Unlike other literary works, when dealing with software infringement, it is almost impossible to make an accurate appraisal of the value of a claim and the potential severity of any infringement. Ultimately it will be a matter of degree in each case and an assessment will have to be attempted having regard to all the circumstances.
Although from a defendant’s perspective the portion of code copied may have made no real qualitative contribution to the defendant’s software product, in the absence of any knowledge about the value of that portion of code, inferences will be drawn from the mere fact that it was taken. If copyright infringement is found, the door swings open to potentially significant damages. The assessment of those damages is without doubt a difficult and imprecise area. Given the legal uncertainties, in practice we find that more and more parties to software infringement disputes are side-stepping a judicial determination of quantum in favour of trying to reach a commercial settlement.
Sarah Hill and Ciara Cullen are Senior Associates in the IP and Technology Group at RPC. Ciara specialises in IP litigation and dispute resolution and has particular experience advising international clients on large scale multi-jurisdictional disputes. Sarah specialises in information technology disputes including claims arising out of failed projects as well as disputes relating to delivery and quality of software.
