Patent World 2003/2004, p. 26-30
Introduction
In 1995 the new Dutch Patent Act introduced a system of Dutch registration patents: these patents are granted on the basis of the application as drafted by the applicant without any examination of the merits of that application. In 2001 the Dutch Ministry of Economic Affairs published an Evaluation of the functioning of this new system. The conclusion was that although most interested parties were not enthusiastic about the new law, no real problems surfaced so that there was no reason for any change.
However, research conducted by Clifford Chance indicates that these Dutch registration patents are used much more aggressively as the basis for litigation than examined patents. In addition, such litigation also seems to be more successful. Given the shaky foundation of these registration patents these facts indicate that this new law is less harmless than expected and may even stifle innovation as opposed to stimulate it.
Dutch Registration Patents
The new Dutch Patent Act of 1995 abolished the prior Dutch system in which patents were examined before grant. That system had been in place under the prior Dutch Patent Act of 1910 and still is in place with regard to European patents as issued by the European Patent Office. It is because of the success of the European system that the Dutch Patent Office saw the number of applications for national Dutch patents drop to such low numbers that it was no longer feasible to support an adequate staff of properly experienced patent examiners.
In response to this development the decision was made not to abolish Dutch national patents altogether, but to introduce a system in which national Dutch patents are no longer examined. The thinking was that such national patents might be a good thing for (Dutch) small and medium sized companies and that they might be persuaded to file more patent applications than before if the threshold would be low.
The Patent Act of 1995 introduced two types of registration patents: a 6-year registration patent and a 20-year registration patent. If one opts for a 6-year Dutch registration patent no examination at all is required. The application as filed by the applicant is 'stamped' and becomes a patent. If one opts for a 20-year Dutch registration patent a search report has to be submitted and will be part of the official file. However, it is still up to the applicant whether he will amend his application because of the results of the search report and he may decide to ignore it completely.
Next to these nationally issued patents one can also obtain a Dutch patent on the basis of a European patent as issued by the European Patent Office after a proper examination of the application. In addition Dutch national patents issued that were issued under the Dutch Patent Act of 1910, which patents were properly examined as well, are also still around for the duration of their remaining 20-year life. These old Dutch patents under the 1910 Act and European patents will be referred to as 'examined patents' as opposed to any patents issued under the Dutch Patent Act of 1995 that are referred to as 'registration patents'.
The Goal of the Research
Bearing in mind that many examined patents are not exact copies of the application as originally filed because the examiner found reasons to restrain the 'enthusiasm' of the applicant, it is obvious that the far-reaching freedom given to the applicant in case of a registration patent is probably too tempting for the average applicant. It seems a bit like telling a kid in a candy store that vegetables are much healthier and simply assume that he will not eat the candy.
This assumption is indeed naïve and it turns out that in a large number of cases pretty damaging prior art does not lead to an amendment of the claims. In addition, The Dutch Bureau of Industrial Property, when asked to advice on the merits if registration patents, does find that a relatively large number of these patents cannot survive and should be revoked in whole or in part.
In case of litigation the result of a novelty search have to be submitted with regard to all registration patents, both 6-year and 20-year patents. This is supposed to provide for an adequate safeguard against abuse of the substantial number of frivolous patents that a system like this produces. However, the burden is totally up to the defendant and the court to evaluate the implications of the result of any search report. The judges will have to decide on novelty and obviousness themselves. Given the fact that having a technical background is no requirement all for being a patent judge, this role does create some additional challenges for the judiciary. They can ask for expert opinions from the Dutch Bureau of Industrial Property or from party appointed experts but is their duty and task to figure out whether the patent is indeed valid, either in whole or in part.
The question that comes up is how this system plays out in real life. One would expect that registration patents do less well in litigation than examined patents, since these patents have not been 'tried and tested' in the hands of a patent examiner. One would therefore expect or hope that courts end up finding in favor of the defendant in registration patent cases more often than in cases on the basis of examined patents. The purpose of this research project was to see whether these expectations matched the actual situation during the period from 1995 until 2003. The results turned out to be very sobering and seem to function as a wake-up call.
Scope of the Research
The District Court at The Hague has sole jurisdiction in patent cases in The Netherlands. All patent decisions by that court are sent to the Dutch Bureau of Industrial property and made available to the public that way. This provides for a unique opportunity to actually examine how patent cases based on registration patents play out compared to litigation on the basis of examined patents. The benefit is that one does not have to rely upon a sample of the population one wants to study, but that one can actually 'map' the entire population.
All decisions of the District Court at The Hague in patent cases in the period from January 1, 19995 through December 31, 2002 were examined. A similarly complete collection of all decisions in patent cases by the Court of Appeals at The Hague is not available. Therefore, the focus was only on decisions in the first instance. However, a random review of decisions by the Court of Appeals does seem to indicate that the higher court does not deviate dramatically from the lower court in terms of how it deals with registration patents or examined patents.
The research focused on the following specific points: (a) are registration patents less often used in litigation than examined patents, and (b) are cases on the basis of registration patents less successful for the plaintiff than cases in which examined patents are at issue.
Registration patents are aggressively used in litigation
How are registration patents used as a weapon against competitors? The number of cases in which the court had to decide upon infringement was counted. Cases or decisions dealing with other issues were disregarded. The latter includes, for instance, cases were the title to the patent was at issue or where the court had to rule on cross border relief and jurisdictional issues related thereto.
In the period 1995 through 2002 we counted 297 judgments in patent cases. 34 cases concerned registration patents. That number corresponds with 11% of all patent cases.
Registration patent were only introduced in 1995. Therefore, it is helpful to see how these cases are spread out over this seven-year period. That is shown in the following picture:
Since registration patents were only available for any applications filed after April 1995, it comes as no surprise that the number of cases is low in the early years. However, in 1999 the number registration patent cases rises to 11 versus 28 cases for examined patents. That means that in 1999 registration patent cases amounted to 28% of all patent cases decided in that year. That is a very high number if one realizes that in 1999 only 6% of all Dutch patents were registration patents.
After 1999 the number of registration patent cases drops dramatically to only 1 case in 2001. It seems that this decline can be attributed to the judgment of the District Court at The Hague in the matter of Koppert v. Boekestein of August 25, 1999. In that case the District Court ruled that there is a strict liability for the owner of a registration patent with regard to any damages caused by enforcing a registration patent that turns out to be invalid. That clearly meant that the owner of a registration patent had to think twice before actually trying to enforce his registration patent. Apparently this message from the District Court had a 'healing effect' and led to a dramatic decline in the number of registration patent cases to almost zero.
However, this judgment was reversed on appeal by a decision of the Court of Appeals at The Hague of September 20, 2001. The Appeals Court argued that the Patent Act does not make a distinction between registration patents and examined patents in terms of liability for wrongful enforcement. Therefore a patent owner is only liable for wrongful enforcement if he knew or reasonably should have known that the patent would be held to be invalid. The fact that after that judgment the number of registration patent cases dramatically rose to 9 in 2002 indicates that this case probably explains the drop in the numbers in the period between 1999 and 2002. In 2002 registration patents stood for 22% of all patent cases, while only 9% of all Dutch patents at the time were registration patents. This also seems to imply that registration patents may take care of 20% to 30% of all patent cases in the coming years.
These numbers do make it clear that registration patents are actively used as a legal weapon. This picture becomes even clearer if one relates the number of registration patent cases to the number of registration patents as a percentage of all Dutch patents that are alive in any given year.
The number of registration patents as a percentage of the total population of live patents in The Netherlands climbs steadily from 1% in 1996 to 9% in 2002. If one compares these numbers with the numbers for registration patent decisions as a percentage of all patent decisions than it becomes clear that the number of registration patent cases is much higher than the number of live registration patents justifies. On average the number for registration patent cases is 2,7x higher than the number of registration patents. In other words: registration patents are used 2,7x more for litigation purposes than examined patents.
It therefore seems that registration patents are not applied for to simply 'collect dust' hanging on the wall of some inventor's office. On the contrary. Even though they have not been examined and the owner apparently did not want to spend too much money to obtain the patent, they are actively used as a weapon in litigation. Bearing in mind that the drop in the number of cases in the period of 1999-2001 is probably a temporary thing, it also seems safe to assume that these high numbers are going to be the trend for the coming years.
Registration patents: more preliminary relief requested
The notion that registration patents are used aggressively is confirmed by the fact that also more preliminary injunctions are applied for on the basis of a registration patent compared with examined patents. Only 44% of examined patent cases are in the form of preliminary proceedings ('kort geding'), while for registration patent the number is 65%.
It does not seem that this high number for registration patents can be explained by the circumstance that registration patents may involve more 'low tech' patents that are easier to digest in litigation. The fact that 50% of the preliminary proceedings leads to an interim judgment, while this only true for 9% of all preliminary proceedings with examined patents seems to contradict such an assumption.
The 50% number seems to indicate that the case was apparently too complicated for preliminary proceedings. The complication is probably in the area of the merits of the patent and its possible scope of protection, where the court is likely to need more advice, either from the parties and their counsel or from experts. Other issues that usually trigger an interim judgment, like matters of evidence and damages, are not that critical in Dutch preliminary proceedings.
This high number of interim judgments in preliminary proceedings seems to indicate that registration patents are not 'ready for trial' and that additional information as to the merits of these patents is required. Proceedings on the merits therefore should be the preferred route, since additional information can be more easily dealt with in such proceedings.
The numbers for proceedings on the merits also underscore this point: only 17% of proceedings on the merits dealing with registration patents require an interim judgment. Strangely enough, the number for examined patents is 45%. However, in proceedings on the merits matters like evidence regarding infringement and the scope of any damages probably also explains why an interim judgment is necessary in such proceedings.
Registration patents are more successful in litigation
Registration patents are used more often and in a more aggressive manner in litigation than examined patents. But it is the final result that matters the most, so the critical question is: how many of these cases are actually won by the patentee?
Given the questionable quality of registration patents one would expect that registration patents are less successful in litigation that examined patents. However, the facts turn out to be different. It turns out that registration patents perform just as well as, and maybe even better than, examined patents.
Examined were all cases, both proceedings on the merits and preliminary proceedings. Counted as an 'infringement' were all decision in which the court found in favor of the patentee, irrespective of whether the case was initiated by the patent owner or whether the case was started by the allegedly infringing party to obtain a declaratory judgment. Counted as 'declined' were all decisions were the patentee lost the case, irrespective of whether the patent was found to be invalid or whether the allegedly infringing activity was found to be outside the scope of the claims.
It turns out that the numbers for registration patents and examined do not differ dramatically. For registration patents we found 30% infringements compared with 25% for examined patents. That does not seem like a significant difference. However, what is striking are the numbers for 'declined': 32% for registration patents against 46% for examined patents. That shows that when the first judgment in a case is a final decision, examined patents score dramatically worse than registration patents: a difference of 50%!
The obvious question that comes up is what the end result of these interim judgments is likely to be. Unfortunately of the cases that were counted as an interim judgment, no final judgment was available, probably because the parties may have subsequently settled the matter. Therefore an educated guess has to be made.
(A) One approach is to assume that the trend that applies to final judgments will also apply for the end result in cases with an interim judgment only. There seem to be no compelling reasons why this would not be a valid approach. Based on this assumption the pictures arises that half of the registration patent cases was successful for the patentee, while this is true for only a third of the cases with examined patents:
(B) Another, more conservative approach, is to apply the trend for cases on examined patents which is a much larger population to all patent cases. In that cases the differences are practically non-existent.
The conclusions on the basis of these 'guestimates' are that registration patents are certainly as successful as examined patents and may actually be even more successful. Naturally one has to bear in mind that the volume of cases on registration patents is still small. However, the number does represent the entire population of these cases and is 'as good as it gets' if one wants to see how registration patents are being treated by the Dutch courts.
It seems to me that both conclusions do give rise to concern. No matter how one appreciates registration patents, it cannot be denied that these patents are of a lesser quality than examined patents. Because the patents have not been examined on novelty or obviousness, the number of registration patents that is invalid must be higher compared with examined patents. In addition, the claims will turn out to be more ambiguous and such should also play out to the detriment of the patentee. This position is supported by the circumstance that in most cases in which the Bureau for Industrial Property is asked for an expert opinion its conclusion is that the registration patent concerned is invalid, either in whole or in part.
Conclusion: change is required
Contrary to the conclusion of the Ministry of Economic Affairs in its Evaluation of the Dutch Patent Act of 1995, this research indicates that the new system of registration patents is not working properly.
Given the marginal quality of these patents, the fact (a) that they are used in litigation much more than examined patents and (b) are at least as successful, if not more, as examined patents, justifies major concerns.
One concern is that any possible defendant that is confronted with a registration patent cannot work on the assumption that since the quality of these patents is low and they have not yet been examined, the burden will be on the plaintiff. The numbers show that underestimating these patents is a big mistake.
It also means that the cost of maintaining the quality of the patent system has shifted from the applicant to the competition. It is no longer the applicant that has to invest substantially to obtain an exclusive right that he can use as a legal weapon against the competition. It is now upon the defendant to try to falsify the claims that may very well be without any real merit right from the start. Whatever costs the defendant he has to make in the process will also be for his account only, without any real possibility to recover a substantial part thereof.
The net result is that such registration patents may very well stifle innovation as opposed to stimulating it. Frivolous claims get rewarded and the moment that one changes what one has been doing, the risk of being confronted with such frivolous patents becomes real. Since it will 'cost a bundle' to make a registration patent 'go away' in litigation, it probably makes more sense to either refrain from what one was doing or simply settle the matter and negotiate a license 'at gun point'.
One course of action would be to altogether abolish this system of registration patents, or nationally issued patents for that matter. If The Netherlands cannot afford to maintain a proper examination capacity at its national patent office, it may be better to do without such national patents. In this regard one also has to bear in mind that the purpose of the system was to enhance the use of the patent system by small and medium sized companies. That goal has not been achieved and these small and medium sized companies probably suffer the most by the present system. The moment they are confronted with a patent claim, they have to hire a patent agent and a lawyer to advise them and the mere cost thereof will be most problematic for exactly these small and medium sized companies.
Assuming that abolishing these nationally issued patents will be 'a bridge too far' for the near future, one should at least amend the system in such a manner that any patentee, who is invoking his patent against a third party will have an obligation to first have a full examination of the patent conducted by the Dutch Bureau of Industrial Property (or the European Patent Office). In that way the costs associated with enforcing the patent will be on the shoulders of the patentee, instead of his competitors. That seems to allocate the costs in a reasonable and fair manner and may substantially limit the possibilities to abuse the patent system by being able to confront people with frivolous patents.
Amsterdam, September 2003