What is precedent and does the EPO have it?

From time to time in the comments of the IPKat there arise debates, sometime heated, about the question of precedent – whether a case, decision or instruction is “binding” on some tribunal or other.  Particularly at the European Patent Office (EPO) the issue is contradictory and confusing.  At first instance there exist the “Guidelines for Examination in the European Patent Office” – what force do these have?  At second instance there is apparently no system of precedent discernible in the black letter law of the European Patent Convention itself, and yet the EPO publishes a book called “Case Law of the Boards of Appeal” at regular intervals, and Boards frequently cite prior decisions of themselves or other Boards.

Where the Boards of Appeal hang out

Frequently to this Kat it seems that the two sides talk across each other, setting up an exaggerated version of a position in order to attack it.

Practitioners from a common law background are treated as seeing a system of precedent where there is none, and not appreciating that decisions of the Boards of Appeal of the EPO, or the Guidelines, are not “binding”.  Practitioners from a civil law background are treated as not seeing the persuasive force of prior decisions and instructions even though they may not be “binding”.

This Kat, in an irenic mood, takes the view that these two apparently opposed positions are not as distinct as they seem, and that the real position is rather more nuanced.  In fact, use of terms that bear the baggage of the legal system in which they are rooted leads to misunderstanding.  Therefore this Kat would like to look at the concept of precedent and see what it means in reality, as opposed to jurisprudential theory.

Binding precedent
In common law jurisdictions such as England and Wales, there is a rule that courts are bound by certain previous court decisions (for present purposes the details of which ones does not matter).  That is, a court is obliged to follow the operative part (ratio decidendi) of a previous decision, in principle even if it considers the previous case was wholly wrongly decided.

Does this mean that judges in London frequently follow cases against their better judgement? Well, no, for several reasons.  First, the majority of cases are not generally seen as having been wrongly decided.  Secondly, there is always the tactic of “distinguishing”, i.e. arguing that the present case differs from the prior case in a materially relevant fact, and that this difference allows a different rule to be applied (this means, to Merpel’s amusement, that calling a case “highly distinguished” usually means the opposite of the compliment that the words might seem to mean).

Practically, the system of binding precedent means that a good way to argue a case under the system is to find a case with analogous facts that has an outcome corresponding to the one that you desire, and then demonstrating the correspondences of your case with it.

Judges do sometimes feel bound to follow decisions that they disagree with, and eventually the only way out is to elevate the case to a court high enough to reverse the precedent.  This happened in the Virgin/Zodiac case, on the question of whether damages are due if a patent is found valid and infringed, but the patent in question is later revoked, where the Supreme Court reversed the rule that had previously applied (criticised by Lord Justice Neuberger [as he then was] and the IPKat here).

Persuasive precedent

Where the Patents Court lurks

Courts whose decisions are not binding may nevertheless be persuasive.  That means that a court is not obliged to follow them, but may choose to do so.  Decisions of the EPO Boards of Appeal can be persuasive in the Patents Court.

What does “persuasive” really mean in practice?  It is important for the present purposes, because it is this model of precedent that this Kat thinks helpful in modelling the EPO system.  Well, first and most obviously, it means that the prior decision is not binding, so the later court can simply choose not to follow it.  Nevertheless a court may be reassured by the earlier finding of a different tribunal, and often will need reasons for diverging from the persuasive decision; as is often the case for UK Courts in relation to EPO Board of Appeal decisions.  This Kat thinks that, in the final analysis, it means that it will assist your case if you can find a suitably analogous case for the subsequent tribunal to follow; and, conversely, that if the analogous case was decided the opposite way from that which will assist you, then you will need to find convincing reasons for the tribunal to take a different line.

No precedent
It seems to this Kat that it is only meaningful to say that there is no precedent in the case that the citation of a prior case will make no material difference, or will be completely ignored.

Caselaw of the Boards of Appeal
Decisions of the Enlarged Board of Appeal
So how then are we to see the “caselaw” of the EPO Boards of Appeal?  Let us consider in turn decisions of the Enlarged Boards of Appeal and then of the Boards of Appeal.  On the face of it, according to the black letter law (and leaving aside petitions for review for the present purpose), an Enlarged Board of Appeal decision only binds a referring Board of Appeal, according to Article 112(3) EPC:

The decision of the Enlarged Board of Appeal referred to in paragraph 1(a) shall be binding on the Board of Appeal in respect of the appeal in question.

However, Article 21 of the Rules of Procedure of the Boards of Appeal give a different story:

Should a Board consider it necessary to deviate from an interpretation or explanation of the Convention contained in an earlier opinion or decision of the Enlarged Board of Appeal, the question shall be referred to the Enlarged Board of Appeal.

This means, in practice, that a Board of Appeal cannot diverge from an Enlarged Board decision without a further reference to the Enlarged Board.

This is stated and arises in a markedly different manner from common law binding precedent, but in practical terms, the outcome is very similar, in that Enlarged Board decisions are pretty much always followed or distinguished (G2/88, on second non-medical use, seems for example to be often “distinguished”).

On the other hand, it does not seem to have been questioned that an Enlarged Board can make an erga omnes statement of law that changes practice universally, applying to all applications and not just the case that is the subject of referral (might might be thought to contravene Article 112(3) EPC, but apparently does not).  For example Decision G 2/08 held that Swiss-type second medical use claims were no longer permissible, not least because their patentability had always been questionable (Reasons 7.1.3). However to ensure legal certainty, and to avoid invalidating many very valuable granted patents, the Board decreed that “the abolition of [the possibility of using the Swiss-type claim format] by the interpretation of the new law given by the Enlarged Board in this decision shall therefore have no retroactive effect”, and set an implementation date of three months from OJ publication before it would apply to future applications. This recognises that the Board’s legal interpretation would be fully binding on all parties to EPO proceedings for any cases filed after the chosen date.

Decisions of the Boards of Appeal
Turning now to the decisions of the Boards of Appeal, again the black letter law suggests a very narrow binding nature of the decision in Article 111(2) EPC:

If the Board of Appeal remits the case for further prosecution to the department whose decision was appealed, that department shall be bound by the ratio decidendi of the Board of Appeal, in so far as the facts are the same. If the decision under appeal was taken by the Receiving Section, the Examining Division shall also be bound by the ratio decidendi of the Board of Appeal.

Once again, however, the Rules of Procedure of the Boards of Appeal paint a different picture, and Article 20 states as follows:

Should a Board consider it necessary to deviate from an interpretation or explanation of the Convention given in an earlier decision of any Board, the grounds for this deviation shall be given, unless such grounds are in accordance with an earlier opinion or decision of the Enlarged Board of Appeal. The President of the European Patent Office shall be informed of the Board's decision.

How is this to be understood?  It is possible to focus on the limitation of this provision – clearly the Board of Appeal can choose to disagree with a previous Board, and “only” has to give its reasons.  But there are several other points to consider.  First, to disagree with an earlier decision, one needs to have read and considered it, so already we are dealing with a system that is more complex than simply a “no precedent” situation:  the earlier decision is not simply irrelevant. Secondly, there must be grounds for the deviation from the prior decision.  The later Board cannot simply choose to take a different path.  So practically, the Board may choose to follow the earlier decision, or depart from it if it has reasons (which it must state).  To this Kat, the criteria are generally analogous to those that are termed “persuasive precedent” in common law systems.

From the point of view of the practitioner, it means that selective and suitable citation of earlier Board of Appeal cases that support the case at hand can assist. Thus practically, the “non precedental” system of the EPO Boards of Appeal ends up being not so different from the “precedental” system of the common law.

There is a hierarchy of decisions of the Boards of Appeal, and they do seem to carry more “weight” depending on their ranking in this hierarchy.  Top of the ranking are those published in the Official Journal of the EPO (albeit that this decision is made by the deciding Board, and not after any independent review). Merpel has noticed however that publication of decisions in the Official Journal has greatly tailed off in recent years.  Have readers noticed this as well, and does anyone know why?

Next are those published in the Caselaw of the Board of Appeal, or the periodic updates, most recently this one (noted by fellow Kat David Brophy here)

Then there is the  distribution indicated on the decision itself, which, in reducing order of importance, is “Chairman and members”, “Chairman”, and then finally “No distribution”.  Also, J decisions (from a Legal Board of Appeal) may be seen to have a gravitas above that of T decisions (from a Technical Board of Appeal), although this may simply be that J decisions are often less fact-specific and therefore of more general applicability.

A very important further consideration is the number of Board of Appeal decisions taking the same line in interpreting a point of law.  A point that is supported by a large number of previous Board of Appeal decisions has more weight than one that is supported only by an isolated case.  When a sufficiently large number of decisions take the same line, it can be elevated to the status of the "established caselaw of the Boards of Appeal", whereupon frequently individual cases are not referred to.  For example, in decision G2/07, where where the Enlarged Board considered the possible effect on third parties if it were to depart from an interpretation given by a lower Board, it was stated:

There can be no "legitimate expectation" that an interpretation of a substantive provision governing patentability given in a decision of the boards of appeal will not be overruled in the future by the Enlarged Board, since recognising such an expectation as legitimate would undermine the function of the Enlarged Board of Appeal. This holds particularly true for issues on which there is no solid body of decisions all to the same effect but where instead the relevant jurisprudence consists only of a very limited number of individual decisions, as is presently the case.

This passage, while confirming that the Enlarged Board can always overrule the interpretation of a lower Board, also confirms that where there is a "solid body of decisions" one can be more confident that a precedental line has been established.  On the other hand, there is less confidence in the precedental value of a "limited number of individual decisions".

National court decisions
It may be worth mentioning as an aside that, while UK courts recognise persuasive precedent in EPO Board of Appeal decisions, the converse is not true. While in the first group of decisions of the Enlarged Board of Appeal it was stated in G5/83 (Reasons, para 6):

The establishment of harmonised patent legislation in the Contracting States must necessarily be accompanied by harmonised interpretation. For this reason, it is incumbent upon the European Patent Office, and particularly its Boards of Appeal, to take into consideration the decisions and expressions of opinion of courts and industrial property offices in the Contracting States.

this view has not prevailed, and in T 0452/91 (as cited in the Case law of the Boards of Appeal; III-H, 3.3) for example it is stated:

The questions of patentability are to be decided solely in accordance with the EPC. No national decision should be cited as if it were binding on the EPO; claims should not be refused by the EPO on the ground that their patentability could not be upheld under the jurisdiction of one member state (point 5.4.1).

At least for issues relating to patent law, it is pretty hopeless to argue on the basis of a national decision.  For wider legal principles, however, national caselaw is sometimes considered relevant, as for example in the recent partiality decision of the Enlarged Board of Appeal in R19/11.

Guidelines for Examination in the EPO

A question related to precedental force of Board of Appeal decisions is that of the legal status of the Guidelines for Examination in the EPO. Several commentators have pointed out that these are “not binding”. Well, yes to a point: in the strict and technical sense, that may be so.  However, they are not without legal force.  First, they are internal administrative instructions to employees from the President of the EPO under Art. 10(2)(a) EPC

The Guidelines themselves state:

The application of the Guidelines to individual European patent applications or patents is the responsibility of the examining staff and they may depart from these instructions in exceptional cases. Nevertheless, as a general rule, parties can expect the EPO to act in accordance with the Guidelines until such time as they – or the relevant legal provisions – are amended. Notices concerning such amendments are published in the Official Journal of the EPO and on the EPO website. 

The reference to “exceptional cases” places the Guidelines, in the opinion of this Kat, high in the “persuasive” category of precedent, so that reference to the Guidelines if they support your case is highly likely to be effective, while achieving at first instance an outcome that conflicts with the Guidelines is likely to be challenging.  While, as David’s post noted, the Guidelines sometime lag behind the caselaw, in general the EPO updates them when a landmark decision emerges.

This Kat was taught as a kitten a distinction, which still seems to hold good in principle, that at first instance an attorney should primarily rely on the Guidelines which are directly aimed at the first instance divisions of the EPO, and that Board of Appeal decisions were not directly relevant except to the extent that they are reflected in the Guidelines. On the other hand, on appeal, it is the Board of Appeal decisions themselves that are relevant, and the Guidelines become of lesser importance, as is reflected by the difference in their treatment in the second paragraph of Article 20 of the Rules of Procedure of the Boards of Appeal:

If, in its decision, a Board gives a different interpretation of the Convention to that provided for in the Guidelines, it shall state the grounds for its action if it considers that this decision will be more readily understood in the light of such grounds.[Emphasis added by IPKat] 

The point is that in contrast to the first part of Article 20 referring to Board of Appeal decisions (see above), reasons for different interpretation do not have to be given.  This reflects the fact that the Guidelines, being "instructions", do not apply to Board of Appeal members when reaching their decisions, reflecting Article 23(3) EPC:

In their decisions the members of the Boards shall not be bound by any instructions and shall comply only with the provisions of this Convention.

This Kat concludes that what is precedental or binding at the EPO is a more nuanced question than it is often treated as being, and that parallels with the common law concept of precedent are not misplaced.  As usual, dear readers, over to you for your views and experiences.

Merpel points out that whether a decision is non-binding and non-influential, non-binding but influential, or binding, is a matter that has a substantial impact on players other than the relevant court or tribunal: it affects the decisions of parties whether to file, oppose, litigate and so on.  It also affects the decisions of legislators as to whether a rule needs changing or not. If it is consistently applied with a consequence that is considered undesirable, legislative intervention is more likely than where bad decisions are distributed among good ones because the latter are not binding.  Therefore, she points out, lack of consistent understanding and terminology between practitioners before the EPO is highly undesirable.

This Kat is very grateful to fellow Kats Jeremy Phillips and David Brophy for extremely helpful comments and contributions in response to the first draft of this post, some of which are incorporated in this final version.