Proportionality – Have We Got It Right in Europe?

The EU Directive on the enforcement of Intellectual Property Rights (IPRs) has been around for over two decades now.  In January 2026, the European Commission (EC) published a study into its application prepared by Ernst & Young (EY).  The study found that courts of the EU “granted permanent injunctions in 94% of patent cases, but proportionality was explicitly assessed in only 0.6% of these cases”. 

The study concludes that such a limited consideration of proportionality is detrimental for innovation. It recommends that the EC issue guidelines to address this.

The study does not explain why more frequent proportionality analysis would be beneficial for innovation. But leaving that question aside, can EY’s statistics really be right? 

Unfortunately, the survey from which they are derived is not public, so one cannot look behind the claim. But we believe that few European patent litigators would recognise those statistics as representing their day-to-day experience. Our own background was defending Nokia against patent infringement claims in Europe when Nokia was the largest mobile phone seller in the world: over those ten years we defended over a hundred patent claims in Europe, and only had one injunction granted against us, which we were able to design around. Is it likely that one defendant avoided injunction in 99% of patent cases when others are being enjoined 94% of the time? And would Europe’s courts really not consider a proportionality defence? 

The study

Directive 2004/48/EC on “the enforcement of intellectual property rights”, known as ‘IPRED’, established a framework of measures, procedures, and remedies for the civil enforcement of IP rights, which has proved largely successful over the 20 or so years it has had effect.

The EC likes to review how things are going and, at the end of January 2026, published its “Follow-up study on the application of the Directive on the Enforcement of Intellectual Property Rights”.  It finds that “A study review of over 600 court decisions across EU revealed that courts granted permanent injunctions in 94% of patent cases, but proportionality was explicitly assessed in only 0.6% of these cases.” (see page 12)

The study then offers findings that are, with respect to the authors, difficult to reconcile. On the one hand, the study states that “national courts differ significantly in the types of IP rights they cover and the extent to which proportionality is considered.” At the same time, the study states that the analysis of national court decisions points only to “small variations in the application of the proportionality principle,” and that “the problem of automatic injunctions exists not only in Germany but throughout Europe.”

Ultimately, the study concludes “the Commission could issue guidelines defining factors which courts should assess when applying the proportionality principle, such as i) the nature of the plaintiff, ii) the economic harm suffered by the parties, and iii) the public interest”

Permanent injunctions in 94% of patent cases?

This statistic seems unlikely and is relatively easy to disprove from public sources. Each year, Trevor Cook of Bird & Bird publishes reviews of the year of UPC patent litigation and English patent litigation in the Journal of Intellectual Property Law & Practice^[1]. He includes tables of cases.

In the United Kingdom, in 2025, 11 patent infringement cases reached trial at first instance. Those trials addressed 14 patents. 4 of those patents were found valid and infringed. 10 were not. 

In 2024 the statistics are rather similar: of 21 patents considered, only 4 were found to be valid and infringed. In 2023 it was 6 out of 22, in 2022, it was 7 out of 27, in 2021 it was 10 out of 32.  One can conclude that the United Kingdom is not granting injunctions in 94% of patent cases.

Europe’s Unified Patent Court (UPC) has not been running for long, but in 2024 it gave judgment in 18 patent infringement cases. The patent owner won in 8 of those cases. In 2025 the UPC decided infringement cases in respect of 55 patents. The patent owner prevailed on 20. 

These figures demonstrate that, even in patent infringement cases that reach a final judgment, only a minority result in a finding of infringement. As this is an essential precondition for the grant of an injunction, we can see that the 94% statistic is clearly not correct. Something has gone wrong in the analysis.

But the above, of course, is only part of the story. Most filed patent infringement cases never get to judgment at all. This is particularly the case in the German regional courts. A German patent trial is unlike a trial that readers from common law countries will be familiar with. The chamber of three judges has read the papers before the parties have arrived and often debated the case amongst themselves. At the outset the Presiding Judge will explain the view of the chamber, so that the advocates can use their time to address areas where the judges believe their case to be weak. This is typical of German efficiency, but it does rather reduce the drama and excitement of the verdict, because the parties get a clear idea from the outset as to whether they are likely to win. 

In these German cases, the party that loses the case is responsible for the fees of the court and the costs of the prevailing party.  Court fees are high in German regional courts and can be hundreds of thousands of Euros. If a patent owner withdraws their claim before judgment, the court fee that they pay, and the proportion of their opponent’s costs that they will be ordered to pay, are only a fraction of those they may be ordered to pay if the judgment is given and goes against them. Therefore, as a patent owner, the sensible decision when they see that the chamber may find against them, is to withdraw the case. They may have other patents that they can assert with better prospects. As a result, the cases that proceed to judgment in German patent courts are those where, after several spoilers as to the result, the patent owner still believes their chances to be strong. The published judgments of the German patent courts consequently tend to be limited to those in which the chamber has found infringement. What we do not see in the judgments are the cases where the chamber is leaning in favour of the defence. Those cases will get withdrawn before judgment. 

Thus, the true measure of success rates in German patent courts is not a survey of judgments. One must also take into account the cases that did not reach judgment, many of which can be considered de facto judgments of non-infringement and no injunction. Juve Patent publishes annual reviews of the numbers of patent cases^[2]. Not all German Courts publish statistics, and those that do publish statistics do not publish win rates. But in 2025, in Germany’s leading patent court, Munich, 187 patent infringement claims were withdrawn before judgment whilst only 80 proceeded to judgment. Even if 94% of these judgments (75 of 80) resulted in a finding of infringement and the grant of an injunction, an honest calculation of the permanent injunctions granted by German Courts as a proportion of cases filed will be much, much lower.

In short, whether we look at the UK patents courts, the UPC, or the Regional Court of Munich, claimants rarely win. They certainly do not get injunctions in 94% of patent cases. 

Proportionality in 0.6% of cases?

Article 3 of IPRED states that Member States (MS) shall “provide for the measures, procedures and remedies necessary to ensure the enforcement of the intellectual property rights covered by this Directive.”  “Those measures, procedures and remedies shall also be effective, proportionate and dissuasive and shall be applied in such a manner as to avoid the creation of barriers to legitimate trade and to provide for safeguards against their abuse.” 

IPRED is a directive, which means it needs to be carried through into national law:

• In the UK (which implemented the directive when it was a member state), injunctions have always been a discretionary remedy, and proportionality has been a defence to injunction since Shelfer’s case in 1895[3]. Proportionality is also the main element in the UK’s “American Cyanamid” test for preliminary injunctions^[4].
• In Europe’s Unified Patent Court, proportionality is written into the agreement and rules of procedure^[5], although you will not find the word “proportionate” used: the term “balance of harm” appears instead. The UPC Court of Appeal has recently given guidance for the proportionality defence under Article 3 of IPRED^[6].
• Section 139(1) Sentence 3 of the German Patent Act provides for courts to deny patent injunctions if enforcement causes disproportionate hardship to the infringer or third parties, codifying previous German case law, such as "Heat Exchanger"^[7]. 

So, proportionality is written into the national law of the main patent jurisdictions. Is it really not being applied?

Considering whether an injunction is “proportionate”

Whether a European Court would consider an injunction proportionate depends very much on circumstances. 

It is easier to see that proportionality would be an issue in the case of medical devices. Of the 8 UPC cases in which the UPC found infringement in 2024, proportionality arose in one: in Meril v Edwards^[8]. The case is about medical stents. The UPC found in that case that it would be disproportionate to enjoin the XL size of Edwards’ medical stents because there was not an alternative supplier for the market. As a result, patients needing that size of stent might suffer if an injunction was granted.  An injunction was sustained for smaller-sized medical stents, where alternative non-infringing devices were available. But from this one case out of eight, we can see that, for the UPC at least, the 0.6% figure must be wrong.

In the area of high tech: mobile phones, laptops, video streaming etc, although these implement many thousands of patented inventions, almost all of those are optional. Even amongst cellular telecommunications standard-essential patents, where use of the standard is mandated by national governments, very little of the standard is truly mandatory. The injunction that Nokia received covered a feature of the UMTS standard that could be removed, and that is what we did. It would have been difficult to argue that an injunction over an optional feature was disproportionate, so we did not raise that argument. 

So, in multi-technology products, it is a rare case where it is truly disproportionate to prevent use of a patent. Economic disproportionality may arise where a particular product is well into production and is a small part of a larger product: another UPC case exempted a component for a particular car model which was so far down its manufacturing process that the defendant was allowed to finalise an order. More often, however, European courts may delay the start of an injunction to allow transition to a non-infringing alternative.

One case that is currently before the English Court illustrates where proportionality is an issue. That case is AMI v Martin Baker^[9]. AMI has a patent covering a mechanical device to hold a pilot’s head in place during ejection from an aircraft. This is a good idea because when leaving an aircraft at high speed the force of the slipstream may break the pilot’s neck. Martin Baker makes ejection seats which have an airbag that holds the pilot’s head. So far, Martin Baker seats have saved 7813 lives^[10]: a number that increases almost daily. AMI seeks an injunction against further sales. 

This, surely, would be the perfect example of a disproportionate injunction. No one would benefit if it were granted. There is no alternative product on the market, and even if there were a product available it would take many years to get it certified for use in aircraft. In the meantime, people will die. But this case, a flagship proportionality case, would not be picked up in the statistics above. It would not be picked up because the trial is about validity and infringement. If the patent is not valid, which it may well not be, the court will never get to consider whether an injunction is proportionate. If proportionality does arise, it will not be at the trial or appear in the judgment from the trial. It will arise at a later hearing.

We don’t know whether EY looked at subsequent hearings. If EY has looked only at judgments from patent trials, it will not have seen the proportionality defence being considered at this later stage. We don’t know if EY considered cases where the court allows time for run-off of existing stock or production, and the switch to a non-infringing alternative. We also don’t know if the survey looked at cases where the court uses the phrase “balance of harm” and does not use the word “proportionate”. It seems that the survey must be missing something, because the examples described above alone constitute more than 0.6%. 

But even if the survey did include all of these things, what the result would show is not that European courts are failing to consider proportionality in relation to final injunctions. It would show that proportionality as a defence isn’t often being raised before them. As in Martin Baker above, a court can only decide the issues that are being raised before it at the time that they are raised. In patent infringement cases with the possibility of a design-around, when there are alternative non-infringing products on the market, or where the technology is not applied in life-or-death situations, it is unlikely that the accused infringer would advance the proportionality defence: they would save court time for stronger defences. 

The statistic is not evidence of a deficiency of the courts, which can be corrected by guidelines. It represents that proportion of patent cases in which the infringer considers an injunction would be truly disproportionate, and which progress to the stage where the infringer raises that argument to the Court, and which progress to the stage where the Court issues a written decision on whether to grant an injunction. No doubt Martin Baker will raise this argument if it needs to, but it is not at that stage now and it may well never get there. 

So where does this get us? 

The study expresses concern that EU courts “granted permanent injunctions in 94% of patent cases, but proportionality was explicitly assessed in only 0.6% of these cases” and proposes that “the Commission could issue guidelines defining factors which courts should assess when applying the proportionality principle, such as i) the nature of the plaintiff, ii) the economic harm suffered by the parties, and iii) the public interest”.

Clearly this is incorrect. Whilst the database used for the survey may be accurate, there has evidently been a significant mistake in the collection or interpretation of the data. The reality is that patent owners lose more often than they win in European Courts. The UK courts, the German Regional Courts, and the UPC (even despite its limited number of cases), have all considered proportionality in patent cases when the defence has been raised. 

There is no basis to conclude that European courts are ignoring the IPRED, or that guidelines are needed.