Forum Selling Abroad – Article by Stefan Bechtold, Jens Frankenreiter & Daniel Klerman

 

From Volume 92, Number 3 (March 2019)
DOWNLOAD PDF


 

Forum Selling Abroad

Stefan Bechtold,[*] Jens Frankenreiter[†] & Daniel Klerman[‡]

Judges decide cases. Do they also try to influence which cases they decide? Plaintiffs “shop” for the most attractive forum, but do judges try to attract cases by “selling” their courts? Some American judges actively try to enlarge their influence by making their courts attractive to plaintiffs, a phenomenon known as “forum selling.” This Article shows that forum selling occurs outside the United States as well, focusing on Germany, a country that is often held up as the paragon of the civil law approach to adjudication. As in the United States, German courts attract cases primarily through the pro-plaintiff manipulation of procedure, including the routine issuance of ex parte injunctions in press cases and refusal to stay patent infringement proceedings when the patent’s validity is challenged in another forum. A critical difference between forum selling in Germany and the United States is that court administrators are more actively involved in Germany. As state officials, German court administrators have the incentive to consider the effect of caseloads on government revenue and the local economy, and they use their power to allocate judges to particular kinds of cases in order to make their courts attractive. They also use their power over promotion, case allocation, and resources to reward judges who succeed in attracting cases. Based on an extensive set of interviews with attorneys, judges, and court officials, this Article describes evidence of forum selling in German patent, press, and antitrust law. It also analyzes how German courts compete internationally with courts of other countries.

TABLE OF CONTENTS

Introduction

I. The German Court System

A. Court Structure

B. Court Administration

C. Jurisdiction and Venue

II. Interview Methodology

III. Forum Selling in German Patent Litigation

A. Introduction

B. How Courts Compete

1. Quality and Predictability

2. Speed

3. Limiting Expert Witnesses

4. Stays Pending Resolution of Nullification Proceedings

C. Motives for Forum Selling

1. Judges

D. Evidence Against Forum Selling

IV. Forum Selling in German Press Law

A. Introduction

B. Forum Shopping

1. Ex Parte Preliminary Injunctions

2. Informal Notice of Likely Decisions

3. Speed

4. Pro-Plaintiff Decisionmaking

5. Attorney’s Fees

6. Quality and Predictability

C. Forum Selling?

1. The Perception of Forum Selling

2. Incentives to Forum Sell

3. Alternative Explanations

V. Forum Selling in German Antitrust Litigation

A. Introduction

B. Forum Selling?

VI. International Forum Selling

A. Antitrust Law

B. Patent and Press Law

VII. Generalizing from the Case Studies

A. Similarities to Forum Selling in the United States

1. Broad Plaintiff Choice of Forum

2. The Importance of Procedure

3. Judicial Efforts to Enhance Forum or Even Judge Shopping

4. Limits of Forum Selling and the Risk of Backlash

B. Differences from Forum Selling in the United States

1. The Importance of Judicial Quality

2. Specialization Facilitates Forum Selling

3. The Importance of Administration

4. Political Economy

Conclusion

 

Introduction

Most judges complain about their busy dockets, but some judges seek the influence and prestige that comes from a higher caseload. In a world where plaintiffs can usually choose the forum, some judges mold procedure and manage cases to entice plaintiffs to file in their court. Just as it is widely accepted that plaintiffs engage in forum shopping, some judges engage in “forum selling.” This competition for cases“forum selling”—has previously been demonstrated in common law countries[1] and arbitration.[2]

This Article shows that forum selling also occurs in civil law systems. In theory, civil law judges are apolitical career civil servants who apply the law in a bureaucratic, anonymous fashion. Nevertheless, forum selling is a reality even in civil law jurisdictions. This Article focuses on Germany, which is often described as the model civil law jurisdiction.[3] Drawing on dozens of interviews with German judges and lawyers conducted specifically for this Article, it shows that forum selling thrives in Germany. This Article pays particular attention to patent law and press law, in which there is good evidence of forum selling within Germany. It also analyzes antitrust law, in which competition is more international and German courts compete (not very successfully) against courts in the Netherlands, the United Kingdom, Finland, and other European countries.[4]

The findings in this Article suggest that forum selling is driven by somewhat different factors in Germany than in the United States. Most importantly, while individual judges are the most important agents of forum selling in the United States, court administrators (including state ministries of justice and the executive committee of each court) play an important role in Germany. Through their power to control judicial careers and to create chambers (groups of three judges) that specialize in particular areas of law, court administrators can make their courts attractive.

Nevertheless, the methods used to implement forum selling in Germany are, in some ways, remarkably similar to those used in the United States. For example, just like federal judges in the Eastern District of Texas, German patent and press law judges seem to compete for litigation mostly by interpreting procedural rules in a pro-plaintiff way, which helps shield decisions from appellate review.

As in the United States, forum selling results from venue rules that give plaintiffs almost complete choice of forum in particular kinds of cases. When plaintiffs can choose to sue in only one or two courts, judges have little to gain by aggressively competing for business, because the amount of litigation they can attract is limited. In contrast, when plaintiffs can file nearly anywhere, an enterprising court can gain a sizable fraction of the entire nation’s litigation in a given subject area. So, for example, just three of the 115 regional courts dominate press law litigation in Germany.

As in the United States, forum selling results in a pro-plaintiff tilt. Because plaintiffs choose the forum, judges and administrators who want more cases must make their courts attractive to plaintiffs. While courts compete for cases partly by enhancing the speed and quality of their proceedings, they also do so by more questionable practices, such as ex parte injunctions (which deny the defendant an opportunity to be heard) and allowing plaintiffs to collect damages on patents that may be invalid.

With the possible exception of press law, forum selling in Germany seems to result in less blatantly pro-plaintiff decisionmaking than in the United States. This may reflect the important role of state-level court administrators, who have incentives to take into account the effect biased judging would have on local industry and their own political careers. Because court competition in Germany seems more benign, the problems it causes are probably best addressed through issue-by-issue reform, such as requiring judges to stay infringement proceedings when it is more likely than not that the patent will be invalidated in another forum. While reform of the broad venue rules that allow competition may have other advantages, it may not be the best way to address the problems caused by forum selling, because tightening venue rules would undermine the benefitssuch as judicial quality, speed, and responsivenessthat competition has fostered. Press law, however, may be an area where venue reform would be appropriate, because the pro-plaintiff bias in this area seems more pervasive and especially problematic, and because it reduces the freedom of the press.

More generally, competition for cases can be both good and bad. If courts compete by being more accurate and faster, the justice system will improve. If courts compete by being pro-plaintiff, the justice system will likely worsen. Competition for cases in Germany seems to have resulted in more of the good effects and fewer of the bad ones. Nevertheless, because plaintiffs generally choose the forum, courts interested in hearing more cases are likely to be tempted to compete by being more pro-plaintiff, with the attendant negative consequences. In some ways the debate over the effects of court competition mirrors the debate over competition for incorporation, where some argue that competition leads to an efficient race to the top producing better law,[5] while others argue that competition leads to a race to the bottom because courts tilt the law to protect the people who choose where to incorporate: management.[6]

Part I provides background on the German court system. Part II discusses how interviewees were selected and how the interviews were conducted. Part III analyzes forum selling in German patent litigation. It argues that competition among courts for patent cases has resulted in improvements in the speed and quality of patent litigation. Nevertheless, competition has also led to problematic practices, such as reluctance to stay infringement suits while other courts determine patent validity. This leads not infrequently to situations where a defendant is barred from selling a product which infringes a patent that is later invalidated. Part IV analyzes forum selling in press law, where judges on the dominant courts seem unusually willing to grant preliminary injunctions ex parte, although this denies defendants a key procedural right: the opportunity to be heard. Part V explores the possibility of forum selling in antitrust. Part VI explores the international dimension of forum selling, in particular, how German courts compete with courts from other European countries. Part VII generalizes from the analysis of particular areas of law and compares forum selling in Germany and the United States.

I.  The German Court System

This Part aims to introduce the reader to the German courts, with special emphasis on issues particularly relevant to this Article. Describing a complex federal system in three pages by necessity involves some simplification and glossing over of regional differences.

A.  Court Structure

The German judiciary comprises five branches. The branch called “ordinary courts” (ordentliche Gerichte) has responsibility for the civil law disputes discussed in this Article.

Like the United States, Germany has both federal and state courts, but the structure is very different. The sixteen German states (Bundesländer) run the lower courts, which are responsible for applying both state and federal law. Federal courts generally only hear appeals from the state courts. There are no federal courts of first instance (trial courts) with general jurisdiction.

The ordinary courts system in most cases has three tiers. The 115 regional courts (Landgerichte) are the courts of first instance for larger civil cases, which are the focus of this Article.[7] The twenty-four regional courts of appeals (Oberlandesgerichte) are responsible for hearing appeals (Berufungen) from decisions by the regional courts.[8] Decisions by the regional courts of appeals are subject to a second appeal (Revision) to the Federal Court of Justice (Bundesgerichtshof).[9]

The judges in German regional courts are organized into chambers, which consist of three judges: a presiding judge (Vorsitzender) and two other judges (Beisitzer).[10] One of the judges in the chamber, acting as reporting judge (Be­richterstatter), assumes responsibility for preparing the case,[11] but regardless of who has primary responsibility, the presiding judge largely determines how a case is handled.[12] Final decisions are taken by majority vote and are issued in the name of the entire chamber, so the reader cannot tell who wrote the opinion. There are no concurring or dissenting opinions.[13] As a result, the typical German judge has no ability or incentive to build an individual reputation through his or her voting behavior. Some chambers have a specialized jurisdiction, such as patent law or press law. Nevertheless, even a specialized chamber may be assigned cases outside its specialization, if it would not otherwise have a sufficient caseload.[14]

The regional courts of appeals ordinarily hear only appeals involving more than €600,[15] while the Federal Court of Justice, in principle, hears only appeals if the regional court of appeals allows its decision to be appealed (Revisionszulassung).[16] However, if the value of the appeal exceeds €20,000, an aggrieved party can file a motion with the Federal Court of Justice asking it to review a regional court of appeals decision denying appeal (Nichtzulassungsbeschwerde).[17]

B.  Court Administration

Judges involved in court administration and officials at the state ministries of justice have significant power to determine whether a court will be an attractive venue. This group of people, which we refer to as the “court administration,” consists of two principal groups: the state ministries of justice and court executive committees (Präsidium).[18]

Each German state has a ministry of justice, whose officials include judges on leave from their judicial duties as well as other personnel.[19] The ministry is responsible for hiring judges and promoting them. Judges are not elected, but are typically appointed to office right after having finished their legal education.[20] The most important criterion that ministries of justice consider when appointing judges is the candidate’s grades on the two state law examinations. German judges enjoy tenure up to a mandatory retirement age of 67,[21] and it is rare for a German judge to take another position after retirement. As a result, German judges enjoy considerable independence, as guaranteed in the German constitution.[22] On the other hand, the government controls promotions, which are more common in a career judiciary like Germany’s. An ambitious judge, therefore, may consider potential ministry of justice reactions when making decisions. As a nineteenth century Prussian minister of justice is reputed to have quipped, “judges can be independent as long as I control their promotion.”[23]

The total number of judges is determined by the state parliament, but the ministry allocates the positions among the courts. The ministry of justice also appoints the president of each court, who has administrative responsibilities, including, as discussed below, service on the executive committee. In addition, the ministry allocates funds among the various courts in the state.

Each court has an executive committee consisting of the president and other judges.[24] Judges on the executive committee, other than the president, are elected by all judges on the relevant court. The executive committee divides the judges into chambers, determines whether chambers will have subject matter specialization, and sets the rules allocating cases among chambers.[25]

C.  Jurisdiction and Venue

Unlike in the United States, in Germany there is no distinction between rules governing venue and those governing the personal jurisdiction of the regional courts. A defendant can almost always be sued at her place of residence or business.[26] In addition, depending on the area of the law and the type of case, a lawsuit might also be brought elsewhere. For example, for tort cases, section 32 of the German Code of Civil Procedure (ZPO) allows a plaintiff to sue in “the court in the jurisdiction of which the tortious act was committed.” Venue rules for the principal legal areas discussed in this Articlepatent, press law, and antitrustwill be discussed in detail below. Although a case can be transferred if the original court has no jurisdiction,[27] there are no rules allowing a case to be transferred solely on the ground of convenience.

II.  Interview Methodology

Our analysis is based on thirty-five semi-structured interviews with fifty judges, court officials, and lawyers in Germany. We chose to primarily rely on interviews for this study because of the limited availability of quantitative data about the German court system.[28] The judges interviewed include trial (first instance) judges in the regional courts and appellate judges in both the regional courts of appeal and the German Federal Court. In order to encourage interviewees to respond candidly, we promised anonymity.

Eighteen of the interviews covered patent law, ten press law, and seven antitrust. Patent law interviews took place between April 2014 and January 2019. Interviews in press and antitrust law were conducted between February 2016 and January 2019. Stefan Bechtold and Jens Frankenreiter conducted all interviews in German. With the exception of two phone interviews (one in patent law and one in antitrust law), all interviews were conducted in person.

We selected our interviewees through snowball sampling.[29] This means that, after contacting a small number of practitioners we knew, we used existing contacts to get in touch with new interviewees, and then asked our interviewees for help in contacting additional interviewees. The advantage of this approach is that we gained access to some of the most important figures in each area of the law. The disadvantage is that our interviews may not be representative of the overall population of judges and attorneys in Germany.

Of the eighteen interviews in patent law, eight were conducted with active and former judges. Two of these interviews involved two judges: a total of ten judges were interviewed. Seven of these judges were active at the time of the interview, two were court officials, and one was a former judge. The judges interviewed included current and former judges from the three major patent courts of first instanceDüsseldorf, Mannheim, and Munich. Five were appellate judges.

In eight additional interviews, we talked to eleven lawyers and patent attorneys with offices in Munich or Düsseldorf. Finally, we conducted one interview with three in-house lawyers at a large engineering company in Germany, and one interview with two (nonjudicial) court officials at one of the three major patent courts. In the other subject matter areas, we talked to fourteen attorneys, two judges, and three lawyers working at corporate entities with significant involvement in litigation in the relevant area of the law.

The interviews were conducted as semi-structured interviews. We did not engage in an open conversation, but structured the interviews around a list of questions that remained largely unchanged throughout the process and across the different areas of the law. During the interviews, we made sure to ask all questions to all interviewees, although we sometimes modified the order. We allowed interviewees to elaborate freely in response to our questions and followed up with related or clarifying questions where it seemed helpful. Most interviews were recorded and then transcribed. In some interviews, we were only able to take notes.

To analyze the interviews, two of the co-authors coded the interviews independently from each other according to a preset coding hierarchy, employing the qualitative research software NVivo.

III.  Forum Selling in German Patent Litigation

A.  Introduction

German courts are Europe’s most important forum for patent disputes and are an important venue for the enforcement of patents worldwide.[30] Some authors estimate that more than 60 percent of all patent infringement cases in Europe are brought in Germany.[31] Patent law also plays a special role in the German legal system, as it is one of the few areas where multinational corporations frequently appear before public courts.

There are twelve regional courts with jurisdiction to hear patent infringement cases in Germany.[32] In principle, venue rules in Germany require a plaintiff to sue in a district with a connection to the dispute. For tort cases, the law allows the plaintiff to sue in the district where the tort was committed.[33] In patent law, this rule is interpreted to allow plaintiffs to file a complaint in any district where an allegedly infringing product is sold.[34] As in the United States, before the Supreme Court’s recent decision in TC Heartland LLC v. Kraft Foods Group Brands LLC,[35] this venue rule means that, for a widely distributed product, plaintiffs can bring a claim in any of the twelve courts with jurisdiction over patent disputes.[36]

As mentioned above,[37] decisions of the regional courts can be appealed to a regional court of appeals. For each of the twelve regional courts with jurisdiction over patent suits, there is a different regional court of appeals. The Federal Court of Justice, which hears appeals from decisions by the regional courts of appeal, also has jurisdiction to hear appeals from decisions by the Federal Patent Court (Bundespatentgericht) in patent nullification proceedings.

Given the freedom plaintiffs enjoy in selecting the forum, it is not surprising to find ample evidence of forum shopping in patent litigation: a starkly uneven distribution of the caseload among the twelve relevant courts. The regional court of Düsseldorf has the highest case numbers, followed by the regional courts of Mannheim and Munich. These three courts together hear 80 to 90 percent of all infringement cases in Germany, with Düsseldorf hearing considerably more cases than Munich.[38] In fact, recent estimates indicate Düsseldorf attracts roughly one-third of all the patent litigation in the entire European Union.[39] Given that travel within Germany is fast and German patent attorneys can and do litigate cases all over the country, the only reasonable explanation for this distribution of cases is that patent owners are strategically selecting the court.[40]

Nevertheless, it is impossible to conclude from the distribution of case numbers alone that there is forum selling, as well as forum shopping. The observed distribution of cases could have resulted from something other than the conscious efforts of judges or other officials.[41] The distribution of cases might prove forum shopping, but it would not prove forum selling, because forum selling is the deliberate attempt by courts to attract more cases.[42] While forum shopping refers to plaintiffs’ strategic choice of forum, forum selling focuses on actions by judges and court administrators to influence their caseloads.

Düsseldorf has been the most important for venue for patent infringement proceedings in Germany since the end of World War II, when most Berlin-based attorneys practicing patent law moved there. Until the early 1990s, the so-called “localization principle” (Lokalisationsprinzip) restricted attorneys to representing clients in a single court. As a result, attorneys usually initiated proceedings in the court in their district. Theoretically, a Düsseldorf-based attorney could have brought a case in Munich as well. However, in order to do so, the attorney would have had to cooperate with an attorney registered with the Munich court. This was particularly difficult because attorneys were barred from working in law firms that included lawyers registered in different districts. Given that many attorneys practicing patent law were based in Düsseldorf, it is not surprising that Düsseldorf had the highest case numbers. Besides Düsseldorf, Munich had a significant share of cases due to the fact that the German Patent Office has been based in Munich since 1949.

Important shifts in case numbers have occurred since the early days of the German Federal Republic. Although the shifts are modest compared to the meteoric rise of the Eastern District of Texas as the dominant venue for patent litigation in the United States, they are significant. Most importantly, Mannheim has emerged as the second most important patent venue, eclipsing Munich.

As will be demonstrated below, while sseldorf’s position partly reflects historical factors, the present distribution of casesas well as the handling of cases at the important patent venuesis the result of forum selling on the part of judges and court officials. Düsseldorf was able to maintain its position as the most important patent venue through a combination of the judges’ patentee-friendly handling of cases and the court administration’s active staffing policy. In addition, Mannheim obtained its status as the second most important venue by positioning itself as a faster alternative to Düsseldorf at a time when Düsseldorf’s success in attracting cases led to substantial delays. This rise, in turn, led Düsseldorf to increase the number of judges focusing on patent law in an effort to increase the speed of its proceedings. Lastly, in recent years the Munich District Court has been fighting its decreasing case numbers by adopting an alternative, more plaintiff-friendly way of handling proceedings in a conscious effort to cater to plaintiffs.

One reason German courts are generally so attractive for patent litigation is that it is relatively easy for patentees to obtain an injunction against infringing parties. Not only are patent infringement proceedings relatively quick[43] and cheap[44] but also, in contrast to the United States and many other jurisdictions, injunctions are generally awarded solely based on a finding of infringement. While patent and antitrust law impose certain constraints on full enforcement of intellectual property rights,[45] courts usually do not give any weight to equitable considerations like the consequences of an injunction for the defendant.[46] Furthermore, the defendant in an infringement suit cannot assert as a defense that the patent is invalid. This is a consequence of the bifurcated system of patent litigation in Germany.[47] Challenges to patent validity can be raised only in administrative proceedings before the issuing authority (either the European Patent Office (EPO) or the German Patent and Trademark Office) or before the German Federal Patent Court (either as a direct challenge or as an appeal against a decision by the German Patent and Trademark Office).[48] Infringement proceedings, by contrast, are heard in the ordinary civil courts,[49] which do not have jurisdiction to decide patent validity. Rather, the defendant challenging the patent’s validity must initiate a separate nullification proceeding in the appropriate patent office or court.[50] If the patent infringement and the patent nullification proceedings overlap, the infringement proceedings are not automatically suspended. Rather, the ordinary civil courts enjoy discretion whether to stay the proceeding until the nullification proceeding is resolved.[51] As nullification proceedings usually take considerably longer than infringement proceedings,[52] unless infringement proceedings are stayed, even holders of patents that are later declared invalid are usually able to obtain enforceable injunctions, which allow them to temporarily bar the alleged infringer from selling its products.[53]

B.  How Courts Compete

Just like in the United States, German patent judges who are interested in attracting litigation to their court have an incentive to cater to the preferences of plaintiffs. Defendants usually have no way to influence the venue in which the case is heard. As mentioned above, if the court selected by the plaintiff has jurisdiction, there are no rules allowing a defendant to request that a case be transferred to another court. In addition, a potential defendant cannot preempt an infringement suit by filing for declaratory judgment.[54]

Plaintiffs focus on two main dimensions when they decide where to bring a case: the quality and predictability of the judgment and the speed of the proceedings. According to patent attorneys, there are considerable differences between the quality and speed of proceedings at different courts. Düsseldorf has a reputation for delivering the highest quality decisions. They are perceived as well reasoned and in accord with the case law of the Federal Court of Justice, and thus unlikely to be challenged or overturned on appeal.[55] This perception varies, however, for different technical areas. In particular, Mannheim has acquired a reputation for being wellversed in cases concerning telecommunications.[56] Mannheim also has a reputation for adjudicating cases more quickly than Düsseldorf.[57] Munich’s reputation on both dimensions has improved in recent years, but it is considered to trail Düsseldorf in quality and Mannheim in speed. By contrast, we found no evidence of systematic differences between courts in the interpretation of the substantive law.

The different strengths of individual courts seem to appeal to different kinds of plaintiffs. Patent attorneys report that they tend to file complex cases in Düsseldorf, while they prefer Mannheim when time is the dominant concern.[58] In individual cases, plaintiffs might also choose to pick a slower court in order to delay the resolution of the case. Mannheim appears to attract many cases involving the telecommunications industry, while most cases involving pharmaceutical and chemical industries are filed in Düsseldorf.[59]

Other differences that plaintiffs take into account when deciding where to file include the handling of preliminary injunctions,[60] the acceptance of English-language documents,[61] and the enforcement of injunctions.

1.  Quality and Predictability

The ability of individual judges to improve the quality of decisions is limited. The most important determinants of the quality of a judgment are the experience and the ability of the sitting judges, and that is relatively uniform across the relevant German courts. Of course, it is possible that some judges work harder, but attorneys generally describe patent judges across all courts as dedicated and highly motivated. One can speculate whether Mannheim’s preference for quick decisions comes at the price of slightly lower quality, while Düsseldorf accepts longer durations in order to ensure a high quality of decisions.

Our interviews indicate, however, that judges invest in increasing the predictability of their decisions. In Germany, whether a decision is published largely depends on a judge’s decision to send it to a database provider or publishing house.[62] Judges in Düsseldorf attempt to publish as many decisions as possible in a conscious effort to increase the predictability of their decisionmaking.

Court administrators also play a key role in ensuring quality. Court administrators have a large influence over judges’ careers. Also, they have considerable influence over judicial workloads, as they decide how many judges at in a court hear patent cases, and whether patent judges also hear cases in different fields of the law. A number of interviewees attributed the perception of Düsseldorf as the highest-quality provider of patent litigation at least partly to the very active policy of the court’s administration to assign excellent young judges to patent cases, to make sure that more senior roles are given to judges with experience in patent law, and to provide patent judges with career incentives that make it attractive both for talented young judges to join this field and for older judges to stay in the field. Other court administrations, by contrast, are perceived not to treat judges specializing in patent litigation favorably in this regard.[63] In particular, for many years the courts in Munich, in keeping with their general approach to judicial assignments, did not allow judges to spend most of their career in patent law. Also, courts can use their power to manage the caseload of patent judges in a conscious effort to provide their judges with the time necessary to produce high-quality judgments in a short time.

2.  Speed

While judges have limited ability to increase the quality of proceedings, they use various tools to speed up proceedings in an effort to appeal to plaintiffs. First, they make use of the fact that they have broad discretion in how they manage proceedings before their court. Judges have a lot of discretion in when and how often they schedule oral hearings and how many rounds of briefs they allow the parties to submit. These decisions have a major impact on the length of proceedings. Interestingly, while the three main patent courts follow different approaches in how they handle their procedures,[64] they all aim to resolve cases quickly. Attorneys even report that judges advertise the speediness of their proceedings, promising to resolve a case in less than a year.

Not only do judges attempt to cater to plaintiffs by increasing the speed of proceedings, but also court administrations across the country have used their power to create additional patent chambers[65] to ensure that the speed of the proceedings does not suffer, despite increasing case numbers. One example of such an effort is the creation of a second patent chamber in Düsseldorf in reaction to Mannheim’s rise during the early 2000s. Düsseldorf, like all courts at that time, had only one dedicated patent chamber. Due to increases in caseloads, the duration of proceedings increased, and attorneys were increasingly looking for alternatives. In that environment, Mannheim succeeded in attracting a considerable number of cases by offering quicker proceedings.[66] Also, patent judges in Mannheim acquired a reputation for the quality of their decisions and for handling cases in a patentee-friendly way. In 2005, Düsseldorf reacted by establishing a second patent chamber, which allowed them to significantly reduce the duration of proceedings.

Increasing the speed of proceedings in most cases favors the plaintiff, who is able to obtain an injunction at an earlier point in time.[67] While we claim that judges and court administrations provide for speedy proceedings in an attempt to cater to plaintiffs, we have no reason to believe that attempts to speed up proceedings favor plaintiffs in an unfair way. Most importantly, we have heard of no reports of instances in which defendants did not have sufficient time to prepare.

3.  Limiting Expert Witnesses

The length of proceedings depends, to a large extent, on whether expert witnesses are brought in by the court. German judges have the ability[68] to appoint expert witnesses to provide an opinion on the technical aspects of a case.[69] Kühnen estimates the involvement of an expert witness causes a delay of about two years.[70]

It seems reasonable to assume that, the more experience a patent judge has, the less often that judge will rely on expert witnesses. Accordingly, Düsseldorf is reported to employ expert witnesses only rarely.[71] Mannheim’s rise is partly attributed to the fact that they also used expert witnesses sparingly.[72] Munich, by contrast, until 2009 had a reputation for higher rates of expert involvement,[73] causing attorneys to counsel their clients against bringing patent suits there. As one attorney put it:

People stopped going to the Munich court because they had the practice of involving expert witnesses in around 80, 90, percent of the cases. . . . Therefore, we just avoided going there. It took too long. . . . For some time, some people called it malpractice to go to Munich. Along those lines, it was that extreme. I am exaggerating only slightly.[74]

The use of expert witnesses by the regional court in Munich arguably was one of the factors contributing to decreasing numbers of patent cases filed there. In 2009, judges at the regional court of Munich announced the introduction of the “New Munich Proceedings” (Neues Münchner Verfahren), a series of changes to their way of dealing with infringement proceedings. Among other things, they announced that they would, in the future, only rely on expert witnesses in rare circumstances.[75] To publicize these changes, the Munich judges gathered a group of about two hundred judges, court officials, industry representatives, trial lawyers, and patent attorneys in a gymnasium of a local school.[76] As demonstrated by the following quote from a judge at another court, these changes are seen as an effort to attract more litigation which was (at least temporarily) successful:

Yes, this was an attempt to use a new way of handling proceedings as a means for advertisement.[77]

A second judge more generally perceived judges in Munich as competing for litigation:

One can in fact not say that judges advertise their court to lawyers. But I believe in Munich this is partly done nowadays. That the presiding judge actively approaches lawyers.[78]

While the current handling of expert witness appointments, at least in one court, seems to be a result of forum selling, we lack a basis to claim that it has led to a procedure that unfairly favors the plaintiff over the defendant. Certainly, as mentioned before, reducing the length of proceedings mainly serves the plaintiffs’ interests. At the same time, we find no evidence that the decision of the judges at the Munich regional court to resolve more cases on their own has led to a decrease in the quality of their decisionmaking.

4.  Stays Pending Resolution of Nullification Proceedings

While the methods of competition discussed so far seem not to have resulted in a decrease in quality (and may even have enhanced it), the reluctance of courts to stay proceedings pending invalidity challenges is much more problematic. As mentioned above, judges in infringement proceedings do not have jurisdiction to rule on the validity of a patent. Instead, patent validity challenges must be brought in separate revocation or nullification proceedings at either the relevant patent office or the German Federal Patent Court.[79] According to section 148 of the German Code of Civil Procedure, judges have the power to stay infringement proceedings until completion of a pending revocation/nullification proceeding. Because nullification proceedings usually take considerably longer than infringement proceedings,[80] a decision to stay a proceeding has a significant impact on the duration of the infringement proceeding.[81] Also, a decision not to grant a stay creates a danger that the plaintiff may obtain an injunction despite the fact that the patent at issue is later declared invalid.[82] The plaintiff, of course, would prefer not to stay the infringement suit in order to both make the infringement suit conclude faster and get the benefit of a favorable judgment of infringement before the patent is declared invalid.

According to case law of the Federal Court of Justice, proceedings should be stayed if the court considers it more likely than not (überwiegend wahrscheinlich) that the patent will be held invalid.[83] In practice, judges at the three most important patent courts have only granted a defendant’s request to stay the case in a small fraction of cases in recent years.[84] For example, Kühnen and Claessen estimate that requests to stay proceedings are granted in only ten percent of all cases.[85] Practitioners report that courts also differ in their willingness to stay proceedings.[86] The restrictive handling of requests to stay proceedings is communicated openly by the courts in a way that seems to be at odds with the case law of the Federal Court of Justice. For example, the regional court in Munich published a set of instructions called “Information concerning the Munich procedure in patent proceedings” (Hinweise zum Münchner Verfahren in Patentstreitsachen), which state with regard to stays:  

The chamber handing down the decision will . . . normally only in exceptional cases come to the conclusion that there is a sufficient likelihood that the patent in question will be invalidated (or substantially confined) in the validity proceeding and stay the proceeding.[87]

A newspaper article cited a judge from the Munich court as stating that infringement proceedings would only be stayed if there was roughly an eighty percent chance that the patent would be declared invalid in the nullification proceeding.[88]

There is no reason to believe that this restrictive practice is justified by the fact that most invalidity proceedings are unsuccessful. Cremers et al. analyzed the outcomes of all cases at the three biggest patent courts in Germany between 2000 and 2008 in which the court issued a judgment despite ongoing nullification proceedings.[89] They found that, among all cases in which the regional court had found the defendant to be at least partly infringing, in 35.5% of the cases the patent was later declared at least partly invalid. In another 45% of the cases, the challenge to the patent validity was withdrawn before decision.[90] This means that, among the cases in which the question of validity was decided, the patent was declared at least partly invalid in almost two thirds. Kühnen and Claessen’s analysis suggests that this problematic situation has continued.[91]

It was not always the case that requests to stay infringement proceedings were handled this way. Until the late 2000s, judges in Munich granted stays at considerably higher rates.[92] This was one of the factors that made Munich an unattractive venue for plaintiffs and led to the introduction of the New Munich Proceeding in 2009. As one judge from another court stated:

The perception of Munich is that [the introduction of the New Munich Proceedings] has led to a decrease in the duration of proceedings. Though before the change the duration of proceedings had been excessive compared with other courts. Proceedings were stayed often in case of parallel nullification proceedings. And I have the impression from talking to many attorneys that the perception is that it has changed to the better.[93]

Besides announcing that they would appoint expert witnesses only rarely, the judges pledged to change their approach to motions to stay.[94] Reportedly, this change has contributed to luring plaintiffs back to Munich. One lawyer commented on this change as follows:

But in practice there were significant differences. One had the impression that in Munich they basically had the view: Well, if we are not absolutely sure about the question of infringement, we can do away with the need to obtain evidence or appoint an expert witness if the patent is invalidated, so we would rather wait. That was in the past. This in fact led to us avoiding Munich, because one had to take into account the fact that the risk of a stay was high. This again led to a counter reaction, with the presiding judge in Munich saying, In Munich, we do not stay proceedings, which is also bizarre.[95]

These days, judges seem to be well aware that a more generous handling of requests to stay the procedure would potentially lead to plaintiffs avoiding their court. One attorney quoted a judge at one of the major patent courts as saying in relation to a number of cases which had been stayed at this court:

Well, I do not really dare presenting these results, because it might potentially threaten our reputation as a patentee-friendly court if I tell you that during the last year we stayed one third of the cases in which this issue played a role.[96]

Like many procedural devices used by the Eastern District of Texas to attract patent litigation,[97] the decision by a German court whether to stay a proceeding is largely shielded from review by the Federal Court of Justice. Section 148 of the German Code of Civil Procedure grants judges discretion regarding their decision whether to stay a procedure.[98] There seems to be no case in which the Federal Court of Justice has reversed a decision by a lower court to stay or not to stay a patent infringement proceeding.

Some commentators have defended the courts’ restrictive approach. They mostly argue that a stay of the procedure would de facto shorten the validity period of a patent.[99] Besides, they point to section 717(2) of the German Code of Civil Procedure, which grants alleged infringers a right to claim damages if an injunction is later revoked. Therefore, they suggest that patentees would not risk enforcing their injunctions if they thought their patent vulnerable.[100]

In spite of such arguments, the practice of the courts has come under criticism in recent years, and there are good reasons to believe that the current practice unfairly favors plaintiffs.[101] First, as noted above, empirical studies find that patents that are the subject of both infringement suits and parallel nullification or revocation proceedings are, in fact, often invalid. Second, the prospect of damages is unlikely to deter patent plaintiffs from suing on weak patents, because defendants subject to injunctions later revoked actually do not consistently recover all losses they suffer when they stop selling the allegedly infringing product. Especially when the defendant was about to enter a new market, it will often have a hard time proving its losses.[102] In addition, there are situations in which the defendant cannot sue for damages after the patent is invalidated, but can only recover for unjust enrichment.[103] The evidence therefore suggests that the prospect of damages does not deter plaintiffs from suing to enforce weak patents or using the threat of an injunction as a bargaining chip to extract wide-ranging concessions from the defendant, potentially even including the withdrawal of the defendant’s challenge to the validity of the patent.[104]

In sum, forum selling has contributed to creating and sustaining a situation in which courts seldom stay infringement suits when there are parallel proceedings regarding patent validity. This is problematic, given that Germany’s bifurcated patent litigation system makes it impossible for the defendant to challenge patent validity in an infringement suit. As a result, patent defendants may lose infringement suits based on invalid patents that they were unable to challenge in time. Judges were supposed to protect against this danger by staying infringement proceedings when invalidity was likely, but competitive pressures have rendered this safeguard ineffective.[105]

The fact that German judges are reluctant to stay infringement proceedings while other courts decide validity is similar to the practice of U.S. judges in the Eastern District of Texas seldom staying suits pending reexamination in the U.S. Patent and Trademark Office.[106] The consequences in Germany, however, are much more severe. In the United States, a patent defendant can challenge the validity of the patent in an infringement suit, so refusal to grant a stay does not mean that the defendant lacks the ability to challenge the patent in timely fashion. In contrast, in Germany, the courts that hear infringement cases lack jurisdiction to hear challenges to patent validity so, without a stay, defendants have no ability to challenge the patent before they are found to have infringed it.

C.  Motives for Forum Selling

1.  Judges

As in the United States,[107] forum selling runs counter to common beliefs about how judges behave. In continental Europe, legal theory is still largely dominated by a legalistic view of judicial behavior. According to this view, judges care solely about applying the law correctly, not about policy or caseloads. Even if one were to adopt a more realistic approach, there are ample reasons to believe that judges would not want to attract patent cases to their court. Because patent cases are assigned to specialized chambers, increasing the numbers of patent cases could significantly affect the workload of judges working in patent law without providing them with any direct benefit. More generally, patent law tends to involve complex, highly technical fact patterns, and it can be a formalistic area of the law requiring extensive expertise in specialized case law that is difficult to apply elsewhere.

i.  Interesting Cases

Judges who specialize in patent litigation probably do so because they enjoy working in this field. Patent law is considered intellectually challenging. Also, judges have the opportunity to work with some of the most sophisticated clients and lawyers. Patent litigation often involves multi-million Euro disputes, meaning that parties will often have the financial resources to finance the litigation until it is resolved by courts. This makes patent law an intellectually attractive area of the law for judges, as legal disputes are decided by judges rather than settled. Because of the existence of specialized chambers, a small number of judges can determine how patent cases are handled.[108] As one attorney put it:

It’s an advantage and a disadvantage at the same time. They have to work more, but maybe they have more interesting work. Those working in the chamber, in particular the presiding judges, have the opportunity to decide exactly how to proceed. They like to work in this area. And when we talk to the judges, there certainly is a point where they say, Sure it’s too much work but overall the work is great. Why is it great? Just like for us, in reverse: They like the complexity, and fortunately, they also praise the quality of the attorneys.[109]

A judge who contributes to attracting cases to the court has a considerable chance of being involved in these new cases. In this way, the fact that German judges who hear patent cases generally specialize in that area facilitates forum selling. Because American judges generally do not specialize, if one judge wants to hear more patent cases, it will be difficult for the judge to do so, because even if the judge gains a reputation for resolving cases in a way that patent plaintiffs appreciate, random assignment of cases generally means that even if a plaintiff files in that judge’s district, the case is unlikely to be assigned to the favored judge. The Eastern District of Texas was able to overcome this problem by deviating from random assignment and, essentially, allowing patent plaintiffs to choose the judge.[110] In Germany, no such problematic deviation from general norms is necessary. If the small number of judges in a patent chamber or chambers want to hear more cases, they can guarantee that all future patent cases in that court will go to them.

While increasing the number of patent cases in a particular court could create a crushing workload, this need not occur. First, court administrators can divert non-patent cases that these judges might otherwise have had to hear to other judges. In addition, in the long run, court administrators can add additional judges. Unlike in the United States, where adding judgeships requires cumbersome legislation at the national level, German court administrators have the power to increase the number of judges in particular areas. This, in turn, increases a judge’s chances of staying in patent law and even acquiring a more senior position. Falling case numbers, by contrast, would over time result in a more mixed set of cases or even the reduction of the number of patent law chambers and the reassignment of judges to different legal areas.

ii.  Reputation and Power

Judges do not specialize in patent law only for intellectual reasons. Patent cases also provide them with the chance to wield power in an important field and to build a reputation as a successful and influential judge.[111]

Patent judges get to decide commercially important cases; more senior judges even have the power to shape the development of the law in an area that is seen as important to economic development and technological progress. Attracting a higher number of cases to the court provides a larger platform with which to influence the law. Not only do higher case numbers imply more decisions in absolute terms, but, by attracting more cases to the court, judges also increase their chances of being involved in landmark cases. One attorney reported:

I think judges are absolutely interested in getting big, important cases. Sure this can happen. They are happy when a lawyer approaches them with a big telecommunications case, and, in view of that, they would do everything to be somehow attractive. I believe that at least those patent judges who are in office at the moment everywhere are not the ones to say: I am happy if this huge thing ends up elsewhere. They are crazy about those cases. They want them. This is absolutely clear.[112]

Some patent judges also care about their reputations as powerful and successful figures in patent law. Another lawyer answered a question about judges’ motives for attracting litigation as follows:

Disadvantage: more work. Advantage: enhanced reputation. One gets better known. They are only human. Why does a patent attorney or an attorney take on a landmark case even though there might be more lucrative work? Because one gets better known by doing landmark cases. Why do we want to get better known? Because everyone wants to be famous.[113]

The notion that judges are driven by a desire to increase their reputation is also shown by an episode in which a judge issued a judgment right after the end of the oral hearing.[114] Lawyers who were present at the oral hearing attributed this decision to the fact that several journalists were present. Rankings of the most popular patent judges, which have been created by legal trade journals, also indicate that reputation may matter.[115]

iii.  Career Perspectives

Building or contributing to an important patent venue can boost a judge’s career prospects. Unlike in the United States, German judges do not generally stay in the court to which they were initially appointed. Instead, if they perform well, they can expect to be promoted to a more prestigious post.[116] Increasing patent caseloads can increase a judge’s chances of promotion for several reasons. First, an increase in case numbers might prompt the court administration to create additional senior positions which have to be filled with judges with prior experience in patent law. In addition, as noted above, judges with a greater caseload are more likely to be perceived as influential and successful.

In recent years, the pending introduction of the Unified Patent Court[117] arguably provided additional incentives for judges to position themselves as successful patent law judges.[118] It is expected that judges at the Unified Patent Court will be paid much more than others, and it seems reasonable to assume that some judges might also be attracted by the intellectual task of creating a new international court system.

iv.  Personal Gain and the Local Economy

Judges have no direct monetary incentive to attract litigation to their court. Most importantly, their salary is not connected to the number of cases they decide. Still, successful and influential patent judges might profit indirectly from their status. They might give paid speeches at conferences,[119] and they can publish books,[120] which can generate substantial revenue for judges whose interpretation of the law is considered important. In addition, promotion to presiding judge or a higher court or selection to the Unified Patent Court would increase the judge’s salary. To the extent that attracting a higher caseload contributes to a judge’s reputation and enhances the chances of promotion, the prospect of a higher salary provides an indirect incentive for forum selling.

In contrast to the Eastern District of Texas, we find no indication that German judges’ behavior is driven by a desire to help the local economy. Nevertheless, as noted below, court administrators may be motivated by the prospect that a strong patent court could attract technology companies to the area.

2.  Court Administration

In addition to individual judges, the court administrationmost importantly court executive committees and state justice ministriesplay an important role in positioning “their” patent courts to succeed in competition with other German patent courts. One judge from another court even went so far as ascribing the rise of Mannheim to a conscious decision by its court administrators:

Probably at one point people at the ministry [the state ministry of justice of Baden-Württemberg] and the court said: Why do we not participate in this? . . . This is how I imagine it. And then they consistently chose top-of-the-notch judges. I know some of them; they are really good.[121]

i.  Staffing

Court administrators play a central role in forum selling in patent litigation in Germany. Not only does their staffing policy directly affect the quality of the patent chambers,[122] their decisions also determine whether judges have an incentive to attract more cases to the court. For example, if the court administration does not react to increasing case numbers with an increase in the numbers of judges dealing with patent law, the length of proceedings will likely go up and the attractiveness of the venue will suffer. If the court administration does not provide judges with an opportunity to work in patent law long-term and to be promoted for doing so, judges have less of an incentive to invest in acquiring patent expertise. As one judge put it:

When they begin [working in a patent chamber], they know nothing about patent law. It takes two years for them to find their way around. This is tough. You have to encourage that through chances for promotion. Nobody does this because it is so exciting. Instead they need to see that they can enhance their career. I think this is absolutely key.[123]

The fact that Düsseldorf has maintained its position as the most important venue for patent litigation seems to be to a large degree caused by concerted efforts on the part of the court administration to reward patent judges. While on the other hand, Munich’s failure to attract more cases despite the proximity of the European Patent Office and the German Patent and Trademark Office can be attributed, at least partly, to the failure of court administrators to craft policies that would make it attractive for judges to build a career in patent law.

ii.  Court Revenue

These observations raise the question of why court administrators use their powers to strengthen their court’s reputation as a patent venue. This question is particularly relevant because it seems reasonable to assume that the strengthening of patent chambers comes at a cost: judges in other fields of the law might feel unfairly treated if they do not get the same kind of recognition for their work or if the administration does not provide them with the same resources as patent judges.

Arguably, the most important reason why court administrations might invest in successful patent chambers is the substantial revenue from court fees.[124] In response to a question about the consequences of increases in case numbers for the states, one judge stated:

First of all, they profit from such increases. Profit monetarily. Revenues from court fees are very high in this area of the law. . . . And this is still ignoring further effects. Law firms also have their offices at this court venue, pay taxes, create jobs, and so on. I do think this is of importance economically.[125]

Patent law is often described as the only legal area that generates enough income to pay for a court’s operations. For example, some estimate that the regional court of Düsseldorf generated about €7.5 million in court revenues from patent litigation in 2009 alone, and that the regional court of Munich generated between 2 and 5 million.[126] The revenues from patent and antitrust proceedings in Düsseldorf help fund the entire judicial system in the state of North Rhine-Westphalia.[127] This creates incentives for the court administration to provide patent courts with the human and physical resources they need to be successful and to attract cases.[128]

iii.  Aiding the Local Economy

The Düsseldorf court administration is reportedly well aware that a strong patent court helps generate additional income for the local economy.[129] More generally, being perceived as a strong forum for patent litigation is considered a positive location factor for industry. Unlike the United States, where there is considerable debate about whether strong patent enforcement benefits patent trolls at the expense of productive companies, in Germany there is less concern about trolls,[130] and strong patent enforcement is perceived as beneficial to local manufacturing. As the justice minister of Baden-Württemberg stated during preparations for the Unified Patent Court:

A strong patent venue in Mannheim is beneficial for the economy in this state. In particular, the many small and medium-sized enterprises in Baden-Württemberg stand out due to their enormous innovative potential and their impressive creativity . . . . In order to protect the six million jobs in the state, this intellectual property needs to be protected effectively and reliably.[131]

Similarly, in February 2017, the justice ministry of North Rhine-Westphalia issued a press release with the title “North Rhine-Westphalia’s state capital on its way to the World Patent Court.” The press release quoted the minister of justice:

With the implementation of the Unified Patent Court agreement in Europe, Düsseldorf is reaching for the crown among the patent court venues in the world. The patent chambers of the regional court and the court of appeals of Düsseldorf have more patent cases than all other courts in Europe. Worldwide, only Washington is more popular with companies. If President Donald Trump continues his policy of isolating his country economically, even this could change very soon, and Düsseldorf could advance to the leading patent court in the world.[132]

Furthermore, the press release stressed that, “[w]ith 7,000 patent filings per year, North Rhine-Westphalia [was] one of the most important patent locations in Germany,” and that “[t]he judiciary in North Rhine-Westphalia [had] superbly qualified personnel and offers firms legal certainty in speedy patent proceedings.[133]

iv.  Administrator Prestige

Contributing to a court’s success might increase the prestige of the local administration as well as the personal reputation of key players such as court presidents and politicians. In 2010, for example, the regional court of Düsseldorf moved into a state-of-the-art building where patent proceedings involving companies such as Google, Apple, and Samsung are held.[134] After it was announced that one of the Unified Patent Court’s local divisions would be located in Düsseldorf, the state justice minister praised the state government’s decision to create additional judgeships at the Düsseldorf courts as responsible for the success of these courts on a European scale.[135] In a similar vein, the presiding judge of the Court of Appeals of Düsseldorf praised the international orientation of the first-instance and appeals courts in Düsseldorf.[136]

v.  Attracting a Branch of the Unified Patent Court

In the years leading up to 2014, the introduction of the European Union’s Unified Patent Court[137] provided an additional incentive for states and their court and ministry officials to position their courts as an important patent litigation forum: decisions about the locations of local branches of the Unified Patent Court were expected to take into account patent caseloads in preceding years.[138]

D.  Evidence Against Forum Selling

We also find a number of features that are hard to reconcile with the forum selling hypothesis. For example, the court in Mannheim has a reputation for enforcing strict time limits for preliminary injunctions. They do not ordinarily accept any request filed more than four weeks after the patentee learns about an alleged infringement. Similarly, one of the most senior judges in Düsseldorf’s way of presiding over proceedings has brought him a reputation for treating litigants in a harsh and unfriendly manner.[139]

In our view, these phenomena do not imply that forum selling does not exist. Instead, they indicate that other factors sometimes prevail over the desire to hear more cases. Forum selling is a factor, but certainly not the sole factor in judicial decisionmaking.

IV.  Forum Selling in German Press Law

A.  Introduction

German law accords relatively broad rights to private persons who are subject to press coverage or otherwise appear in the media.[140] Not only is it illegal to disseminate false or libelous information, but also, under certain circumstances, the media is barred from publishing photos or videos of private persons, or from reporting truthfully on an individual’s private life. In addition, in sensitive matters (such as criminal investigations), it is potentially illegal to disclose the identity of persons whose actions are reported.

In press law, plaintiffs can ordinarily sue in any regional court.[141] In practice, plaintiffs nearly always request a preliminary injunction (einstweilige Verfügung),[142] which can be requested from the court in any district where the allegedly infringing publication is sold or otherwise disseminated.[143] This means that plaintiffs can sue a television station that broadcasts its content nationwide before any of the 115 regional courts in Germany.[144] As in patent law, this permissive venue rule is the result of an expansive interpretation of section 32 of the German Code of Civil Procedure.[145]

In practice, press law cases are heavily concentrated in a small number of courts. While exact numbers are not available, information from interviewees and from appeals proceedings suggest that the courts in Berlin, Hamburg, and Cologne together hear more than half of all press cases. Substantial numbers of cases are also handled by the courts in Munich and Frankfurt.[146]

The concentration of press law cases at a small number of courts is not a new phenomenon. One historical reason for this concentration is that, until the early 1990s, attorneys were only allowed to represent clients in the court in their home district.[147] Because a case asserting a plaintiff’s right of reply has to be brought at the publishing house’s place of business, most attorneys who wanted to practice press law chose to settle in cities where major media outlets were located. As a result, courts in major media hubs turned into important venues for press litigation. Nevertheless, with the elimination of rules restricting practice to the lawyer’s home city, lawyers now routinely travel to litigate in courts far from their homes. The modern concentration of press cases in a small number of courts therefore requires a different explanation.

B.  Forum Shopping

Attorneys representing plaintiffs in press law cases today engage in an aggressive form of nationwide forum shopping.[148] With the exception of cases seeking to enforce a plaintiff’s right to reply, lawyers strategically select the forum that appears most likely to achieve the best outcome for their clients. Also, lawyers react swiftly to perceived changes in court procedure: a court that handles cases in a way that is perceived as plaintiff-friendly can rather quickly attract a large number of cases, while courts that become more defendant-friendly risk losing a large share of cases.

Forum shopping is boosted by the fact that, in all regional courts that are major press law venues, press law cases are heard by only one chamber.[149] This means that attorneys can de facto shop for the judges who are going to hear their case. While a chamber is composed of three judges, the presiding judge has a large influence. The fact that there is only one press chamber per court also means that a single judge, the presiding judge, has an outsized influence on the treatment of press law at that court.

Also contributing to forum shopping is the fact that many of the more important cases in press law are argued by a small number of attorneys who specialize in press law. As a result, a few attorneys interact repeatedly with a small number of judges, allowing attorneys to learn how individual judges deal with recurring issues in press law and to predict how outcomes are likely to differ from court to court.

In recent years, forum shopping in press law has been the subject of an intense debate between defenders of the status quo and proponents of venue reform. Those who favor venue reform argue that forum shopping has created a system that unfairly favors plaintiffs,[150] and some commentators even argue that this renders press-law venue rules unconstitutional.[151] In 2009, the Federal Ministry of Justice considered the need for the reform of venue rules and the rules governing ex parte injunctions in press law.[152] During the legislative process, a proposal to restrict venue choice in press law (and in unfair competition more generally) was dropped. Still, the Federal Parliament, when voting on the final legislative proposal,[153] expressly requested the government to reconsider restricting venue choice in press law cases.[154] A similar request was made by the state ministers of justice in 2016.[155] It remains to be seen whether these initiatives will result in legislative action.[156]

Several factors are particularly important to lawyers in deciding where to bring a press case, including the willingness of judges to issue preliminary injunctions ex parte, the ability to receive non-binding rulings, the speed of the court’s decisionmaking, and the court’s general pro-plaintiff tendencies. Because of the plaintiffs’ ability to select the venue from among all regional courts in Germany, it is not surprising that the courts that attract large numbers of cases appear to be particularly plaintiff friendly on all or most of these dimensions.

1.  Ex Parte Preliminary Injunctions

One of the most disturbing aspects of German press law—and arguably one of the most troubling consequences of the flexible venue rulesis the willingness of judges at major press law chambers to grant preliminary injunctions without giving the defendant an opportunity to be heard. This practice recently caught the attention of the Federal Constitutional Court, which in two 2018 decisions found that aspects of this practice violated the constitutional rights of the defendant.[157] Before these decisions, the defendant often had no ability to inform the judges of reasons the injunction should not be issued. Because many press cases are time sensitive, the ability to challenge or appeal a preliminary injunction after it was issued does not provide a meaningful remedy. 

The recent decisions of the Federal Constitutional Court confirm the analysis in this section, nearly all of which was written before those decisions came out. The Federal Constitutional Court agreed that the practice of the dominant press law courts was unfair and contrary to the law. While the Court did not try to explain why some German courts had adopted an unfair and unjustified practice, as discussed below, the most plausible explanation is competition for cases or forum selling.

Individuals mostly enforce their rights by asking a court to issue an injunction against a media outlet that is about to publish or has published content infringing her rights.[158] Section 937(2) of the German Code of Civil Procedure ordinarily requires judges to hold an oral hearing before issuing a preliminary injunction. Only in urgent cases (in dringenden Fällen) can judges dispense with a hearing or any other opportunity to respond.[159] Although section 937(2) envisions ex parte injunctions as exceptional, major press law venues routinely decided cases without holding oral hearings. Judges could give the defendant the opportunity to respond in writing, telephonically, or in some other way.[160] Instead, at least until the intervention by the Federal Constitutional Court, after the plaintiff filed for a preliminary injunction, the plaintiff and the judge often discussed the issues on the phone without the defendant being involved, and the judge then decided whether to issue the requested injunction. Even when judges took several weeks to issue an injunction, which would seem to provide ample time for the defendant to be heard, some courts did not give the defendant a hearing or other opportunity to present evidence or legal argument.[161] Although ex parte communication between judges and lawyers is neither uncommon nor prohibited in Germany, deciding injunctions without input from the party to be enjoined is clearly problematic, especially when there was time to do so.

Plaintiffs have a strong preference for quick proceedings, and the time required to decide a motion for a preliminary injunction depends largely on a judge’s decision whether to hold an oral hearing. Therefore, the ability to obtain a preliminary injunction without an oral hearing is one of the most important factors for plaintiffs in deciding where to bring a case. As one lawyer stated:

Those chambers, those courts that I just mentioned, they decide without holding an oral hearing. This is extremely important. . . . For this reason, these courts are absolutely superior.[162]

Another indication of the importance of ex parte injunctions is the rise of Cologne as an important venue for press litigation. Until 2002, judges in Cologne usually issued preliminary injunctions only after holding an oral hearing. During that time, Cologne heard only a small number of press cases. When a new presiding judge took office, she changed this practice and started regularly issuing preliminary injunctions without giving the defendant the opportunity to respond. As a consequence, the number of cases increased substantially.[163] In contrast, interviewees report that other courts still routinely hold oral proceedings.[164] This may explain why these courts are shunned by experienced plaintiffs’ lawyers in press cases. Overall, it seems reasonable to conclude that forum shopping was, until very recently, why most press-law preliminary injunctions were issued by judges who routinely dispensed with the requirement that they hold an oral hearing.

The fact that, as a result of pervasive forum shopping, defendants in many cases had no genuine opportunity to respond to a motion for preliminary injunction is problematic.[165] The opportunity to be heard prior to adverse judicial action is one of the most basic procedural rights[166] and is one of the principal safeguards against unjust or ill-informed decisions.[167] In fact, in September 2018, the German Federal Constitutional Court decided two constitutional complaints challenging the practice of the regional courts in Hamburg and Cologne, which issued preliminary injunctions in press law cases without giving the defendant the opportunity to respond.[168] In a move that at least partly outlaws this practice, the Federal Constitutional Court ruled that courts in principle must give defendants an opportunity to respond to motions for preliminary injunction even if they dispense with an oral hearing.[169] In addition, defendants must be informed about ex parte communication between judges and the plaintiff’s lawyers.[170]

While even before the ruling by the German Constitutional Court, defendants had some opportunities to prevent issuance of preliminary injunctions or to challenge them afterwards, these mechanisms were inadequate. The defendant, for example, can try to prevent an unjust ruling by filing a protection letter (Schutzschrift) before the preliminary injunction is issued. These letters explain the defendant’s view of the facts and the law and are filed in a central registry.[171] All courts that receive a motion for preliminary injunction are required to consult the registry to see whether a protection letter has been filed, and judges must take the arguments in the letter into account when considering whether to grant the preliminary injunction.[172]

Nevertheless, the ability of a defendant to file a protection letter provides only a partial remedy to the problems posed by ex parte preliminary injunctions.[173] First, the defendant may not even know that the plaintiff is likely to seek a preliminary injunction and therefore may not realize the need to file a protection letter. In most cases, a defendant will only learn about a plaintiff’s intent to request a preliminary injunction through a cease and desist letter. While some courts generally accept requests for preliminary injunctions only after such a letter is filed, other courts do not consider such a letter to be a pre-requisite to issuing a preliminary injunction. By not requiring a cease and desist letter, those courts make it difficult for the defendant to anticipate when a preliminary injunction will be requested and thus make it nearly impossible for defendants to inform the judge of their side of the case through a protection letter. For this reason, one commentator argued that it is generally illegal to issue an ex parte preliminary injunction when the defendant has not been sent a cease and desist letter.[174]

Second, the defendant’s ability to file a protection letter may not allow the opportunity to present a persuasive defense because the defendant may not know the arguments put forward in the plaintiff’s motion. Therefore, the defendant might not be able to address and respond to all relevant arguments. In cases where the plaintiff sends a cease and desist letter prior to filing for a preliminary injunction, the defendant may be in a better position because such letters usually contain at least a rough outline of the plaintiff’s arguments. Nevertheless, such letters are not an adequate substitute, because they may not contain all of the plaintiff’s arguments and substantiating evidence. In addition, as mentioned above, the defendant might not receive a cease and desist letter prior to issuance of the preliminary injunction.

The law also offers defendants a number of ways to ask the same court or higher courts to review a preliminary injunction. If a judge issues a preliminary injunction ex parte, the defendant can challenge it later and request a hearing in front of the same judge.[175] Preliminary injunctions can then be appealed to the regional court of appeals, but not to the Federal Court of Justice.[176] In addition, the defendant can force the plaintiff to have the injunction confirmed in principal proceedings, which would make the injunction appealable to the Federal Court of Justice.[177]

Like the ability to file a protection letter, these remedies do not fix the problems posed by ex parte preliminary injunctions. The most important reason is the time-sensitive nature of press law cases. Media outlets are seldom interested in publishing after the time prohibited by the preliminary injunction, unless the case involves a disputed legal issue that might be important in future cases. Defendants therefore rarely appeal a preliminary injunction or force a plaintiff to request confirmation of the injunction in principal proceedings.

In addition, the opportunity to request an oral hearing after the preliminary injunction has been issued does not offer an effective remedy, in part because the oral hearing would ordinarily be held before the same judge who issued the ex parte preliminary injunction. Judges who grant preliminary injunctions are naturally reluctant to overturn their own decisions. Interviewees noted that at least some judges are reluctant to change their mind unless the hearing reveals that the plaintiff concealed major facts of the case. Moreover, if an oral hearing is held after an injunction has been issued, appellate courts will ordinarily not review the prior decision to grant the preliminary injunction ex parte. Any violation of the right to be heard is considered to be “cured” by the later oral hearing.[178]

Finally, the law attempts to protect a defendant by making the plaintiff strictly liable for any damages caused by a preliminary injunction that is later revoked, and by giving the judge the ability to require the plaintiff to post a bond.[179] For press law defendants, these rules offer no meaningful protection. Because it is extremely difficult to measure the harm caused by enforcing a preliminary injunction, defendants almost never sue for damages after the preliminary injunction is rescinded, and judges usually do not require the plaintiff to post a bond.[180]

2.  Informal Notice of Likely Decisions

As noted above, until very recently, judges at all major press law venues usually decided whether to grant preliminary injunctions after discussing the case informally with the plaintiff’s lawyers, often by telephone. This opened up the possibility of a particularly aggressive kind of forum shopping. A judge who was unconvinced by the plaintiff’s arguments would often inform the lawyer of her concerns and offer the lawyer the opportunity to withdraw the motion. The plaintiff might then file for a preliminary injunction in another court.[181] This procedure essentially allowed the plaintiff to shop for the court most likely to grant a preliminary injunction.

3.  Speed

As noted above, press cases are often time-sensitive, so speed matters for plaintiffs. Plaintiffs seek preliminary injunctions to prevent the publication or distribution of particular content. Naturally, for a preliminary injunction to have any effect, it must be obtained before the allegedly illegal content has been broadly distributed. All major venues in press law are willing and able to issue preliminary injunctions within a couple of days.[182] In some cases, plaintiffs have been able to obtain a preliminary injunction against a TV station only a few hours after the TV station announced its intent to broadcast a report later the same day.

Judges at major press law venues take active measures to ensure their ability to act quickly on requests for preliminary injunctions. Unlike many other judges, judges in press law make sure to be at court during normal working hours. German judges enjoy wide latitude as to when and where to work, as long as they do not have to attend oral hearings or chamber meetings. As a result, many judges frequently work at home, where they may be unable to respond in timely fashion to issues that arise. Judges in press law, however, make sure to be reachable at the courthouse during business hours. Attorneys sometimes call judges prior to filing for a preliminary injunction to discuss whether the judges will be available to decide the case within the desired time period. The speediness of a preliminary injunction also depends on court staff being present to issue enforceable copies of the injunction.

More experienced judges are also able to decide cases more quickly. While plaintiffs prefer experienced judges for many reasons, including greater predictability, it usually takes less time and effort for the attorney to obtain a decision from an experienced judge. Inexperienced judges usually need time to acquaint themselves with the relevant case law, and attorneys report that they have to spend more time and effort arguing their case. As one attorney put it:

I am sold down the river if I go to a court which does not do press law at all, which has not internalized the complex case law of the Federal Court of Justice. . . . There are lots of topics with which one has to be familiar. In particular, in a preliminary proceeding in which, in order to effectively enforce personality rights, I want to obtain quick protection for the person affected, sometimes within hours, I need judges who know the topic.[183]

The desire for experienced judges gives an advantage to courts that already have large caseloads. Nevertheless, the emergence of Cologne as a major press venue after 2002 suggests that courts that previously did not have a large caseload can overcome that disadvantage by making themselves attractive in other ways.

4.  Pro-Plaintiff Decisionmaking

Plaintiffs also prefer judges who interpret substantive law in a way that favors the plaintiff. Many decisions in press law turn on a balancing of interests. Courts must decide, in a particular case, between freedom of the press and the privacy and reputation of individuals. Attorneys perceive some judges as being more plaintiff-friendly than others in that they give more weight to the interests of the individuals involved. In the words of one attorney:

In press law, there are a couple of courts that have a reputation for their specialization in press law. And among these there are some courts that have a reputation for being particularly friendly towards the affected individual. . . . [T]here are two courts, Berlin and Hamburg, . . . which have a reputation for being pro-plaintiff, because if I want to sue a media outlet as a plaintiff, then it is classic thinking of attorneys that I go to Berlin or to Hamburg because of their reputation for being plaintiff friendly . . . .[184]

Note that judges are not necessarily perceived as equally plaintiff friendly in all kinds of cases. For example, a judge might be particularly solicitous of plaintiffs in cases involving reports on criminal investigations,[185] or might be prone to awarding particularly high amounts of damages. Nevertheless, all judges at chambers that hear a large number of press law cases are overall perceived as plaintiff friendly.[186]

Uwe Jürgens has shown that decisions decided by major press law chambers are reversed by the Federal Court of Justice at unusually high rates, and that defendants initiate a higher number of successful appeals than plaintiffs.[187] This difference in reversal rates suggests a pro-plaintiff bias in the lower court.

Although the Federal Court of Justice has the ability to correct some of the decisions of the major press law courts, it is not able to fully counteract lower court biases.[188] The most important reason, as discussed above, is that many press cases are time sensitive, so defendants often have no incentive to pursue the lengthy process of appeal to the highest court. Another reason is that the Federal Court of Justice decides only legal issues and must defer to the facts found by the lower courts. Finally, the amount in controversy requirement for appellate review makes some press cases unreviewable.[189]

5.  Attorneys Fees

In some cases, it can be attractive for attorneys to file the case at a court that is known for setting the amount in controversy relatively high, because attorneys representing plaintiffs in minor cases are often not paid by the hour. Instead, they earn the statutory fee, which is paid by the defendant if the plaintiff wins. The size of this fee depends on the amount in controversy.[190] In press cases, where damages are subjective and uncertain, the judge has considerable discretion in determining the amount in controversy.

6.  Quality and Predictability

As in patent law, the quality of decisions matters, too. The regional court in Hamburg, while being perceived as somewhat slower than other courts, is known for its well-reasoned and consistent judgments. Plaintiffs also prefer judges with substantial experience, whose decisions, in addition to being quicker, are more predictable.[191]

C.  Forum Selling?

There are a number of facts indicating that judges and court officials adjust their handling of cases in order to increase the number of press cases brought before their court. That is, in addition to forum shopping, there may also be forum selling.[192] The most important facts pointing to forum selling are the troubling practices discussed in the prior section. Would judges be so eager to issue preliminary injunctions ex parte if they were not actively trying to attract more cases? Would they not at least require plaintiffs to deliver cease and desist letters before (or at least simultaneously with) requesting a preliminary injunction? Would they not try to involve defendants in informal discussions of preliminary injunctions or at least inform them of the issues? These, and other procedures that are in no way mandated by law or precedent but that seem biased in the plaintiff’s direction, are hard to explain as anything other than attempts to attract more cases.

In addition, some lawyers perceive judges and court administrators as trying to attract cases. Motives to do so include the fact that press cases are interesting to judges but not time-consuming, bring publicity to the court, and generate fee revenue.

1.  The Perception of Forum Selling

Some interviewees opined that judges and court administrators, particularly in the regional courts in Cologne and Frankfurt, are deliberately trying to attract more cases. In the words of one attorney:

I have a feeling that it could be like that in Cologne . . . , because it is trying to position itself as a press law venue . . . and then try to attract cases, because a quick reactionone just has to increase the amount in controversy, and it becomes more attractive. . . . And at the moment Frankfurt, they are coming, one has the feeling that they also sort of enjoy these cases.[193]

Another attorney stated:

My impression is that Cologne discovered that as a business model. . . . I do believe that the regional court in Cologne realized that there was a lot of money behind this. Because as a judge or a court, one can quickly make money.[194]

2.  Incentives to Forum Sell

It seems reasonable to infer that both individual judges and court administrators have an incentive to increase the number of press law cases filed at their court. There are a number of reasons why press law is an attractive field for judges to work in. First, it is considered an interesting area of the law, and judges seem to like working in this field. As in patent law, while an increase in case numbers will lead to an increase in a judge’s workload in the short run, the court administration has the ability to divert nonpress cases to other judges in order to avoid congestion in the press chamber. This means that a judge who successfully attracts more press law cases will, in the long run, be able to hear more press law cases and fewer cases from other fields. Second, unlike in patent law, cases in press law are relatively uniform and do not require complex fact-finding. This implies that judges might be able to decide large numbers of cases without strain. This, in turn, is important because judges are evaluated and promoted partly based on the number of cases they dispose of. In the words of one attorney:

I have the impression that press chambers are highly attractive for judges as a place to work. It is an interesting subject. These are quick procedures, many procedures. This leads to a high number of closed cases in the department. For starters, this is attractive for judges.[195]

There are also some indications that court administrators are interested in attracting more press law cases to their courts, perhaps for the media attention such cases bring. One attorney reported being approached by the former president of a regional court of appeals, who asked about the reputation of the local courts in press law and about the reasons plaintiffs chose one court over the other:

[I]f a court has a recognized press chamber, it gains reputation. This leads to a certain name recognition, because the press, of course, loves to write about cases concerning itself. . . . This means that this is attractive in this sense, and I also seem to understand from occasional statements that the presidents of the courts are interested in this. I have heard something like that very clearly from the former president of the court of appeals, who casually inquired about the reputation of the regional court and the court of appeals and who wanted to know why people go [here or elsewhere]. This was very clear indeed.[196]

As is apparent from this quote, courts’ interest in press law may be tied to concerns about reputation. Court administrators might also be interested in press cases because of the revenue they generate from court fees. Although the amount in controversy is usually much lower than, for example, in patent law, substantial revenue may still be generated because of the high number of cases and the relative ease with which they can be decided.[197]

3.  Alternative Explanations

Nevertheless, there is no conclusive evidence of forum selling in press law. The observed phenomena are compatible with the hypothesis that judges differ in their views about the correct interpretation of the law and about the proper handling of press cases. Considerations related to caseloads may play little or no role. In most of our interviews, when asked directly, our interviewees asserted that judges did not act strategically. In addition, examples of courts that took a more defendant-friendly approach towards handling cases after a change in the composition of the chamber responsible for press law strongly suggest that, if forum selling exists, it is not a universal phenomenon.

Even if judges are not actively seeking more cases, the broad venue rules which enable nationwide forum shopping are problematic. These rules allow plaintiffs to litigate in the most plaintiff-friendly courts in the nation. This means that a handful of judges and courts decide the overwhelming majority of press law cases and do so in a way that seems excessively pro-plaintiff. Even if the variation in judicial treatment of press law cases were motivated entirely by good-faith interpretation of the relevant laws and precedents, it is troubling that venue rules empower the judges who happen to take the most pro-plaintiff positions and that defendants must conform to their views. Even more worrisome, many important German press courts seem to have moved to an equilibrium of granting ex-parte injunctions and denying defendants their right to be hearda situation that became so problematic that the German Federal Constitutional Court had to intervene in the fall of 2018. This is especially troublesome in press law, where a pro-plaintiff position is one that limits the public’s access to information. While the recent decisions by the Constitutional Court will likely lead to a decrease in ex-parte injunctions, it remains to be seen how this will affect competition between courts in press law.

V.  Forum Selling in German Antitrust Litigation

A.  Introduction

While U.S. and European antitrust law generally share common goals, until recently they have relied on different enforcement mechanisms. American antitrust law has traditionally relied on a complex mix of private lawsuits and public enforcement by the Department of Justice and the Federal Trade Commission. In both the European Union and in its individual member states, antitrust laws have traditionally been enforced solely by public antitrust agencies. Enforcement by private parties only became a significant part of European antitrust law after a decision of the European Court of Justice in the 2001,[198] the European Union Council Regulation 1/2003,[199] and a European Commission white paper in 2008.[200]

The typical way private parties enforce the prohibition against cartels[201] is to file a follow-on damage lawsuit. Once a European antitrust authority has determined that companies have engaged in an illegal cartel, a customer of a cartel member can file a lawsuit claiming damage for supracompetitive prices the customer had to pay for goods produced by the cartel.[202]

Within Germany, plaintiffs in such follow-on lawsuits have considerable leeway in deciding where to file their lawsuit. Twenty-four regional courts have jurisdiction to hear antitrust follow-on damage cases.[203] The cartel member may be sued at its seat of incorporation or where the tort was committed.[204] The phrase “where the tort was committed” is commonly interpreted not only to cover the place where the tortfeasor acted, but also where the protected legal interest (i.e. competition) was harmed. In follow-on damage lawsuits concerning nation-wide cartels, that means the plaintiff can usually sue in any court district in which the overpriced product was sold. This often allows plaintiffs to sue in any of the twenty-four courts with subject matter jurisdiction.

The regional courts typically have a specialized chamber for antitrust cases (Frankfurt has two chambers). This not only allows judges to develop expertise, but it also enables plaintiffs to predict with some certainty which judges will hear their case. Even if the judge with primary responsibility is chosen randomly from among the judges in the chamber,[205] the presiding judge of the chamber has a large influence. As a result, the plaintiff can predict the judge who will have the largest sway over the case with certainty (by suing in a court with one antitrust chamber) or fifty percent probability (by suing in a court with two antitrust chambers).

B.  Forum Selling?

Our interviewees consistently assert that plaintiffs strategically choose the venue to file their follow-on damage lawsuits. Important factors driving this decision are whether the court handles its follow-on proceedings in a speedy and effective manner. The experience of the court handling such cases is also important.

The main challenge for plaintiffs in antitrust follow-on damage lawsuits is satisfying its evidentiary burden. According to general principles of German civil procedure, the plaintiff must prove: (1) that a cartel existed; (2) the particular cartel in question caused prices to increase in general; (3) the cartel had a negative impact on the plaintiff in particular; and (4) the precise damages caused by the cartel.

Proving the existence of the cartel is not difficult for the plaintiff, as the German Cartel Office or the European Commission has usually determined this element in their previous public proceeding against the cartel, and their determinations are binding on the courts.[206] Proving the other three elements is challenging. Given the limited information the plaintiff usually has and the restrictive discovery afforded by German courts, it is nearly impossible for the plaintiff to provide sufficient evidence.

Regional courts in Germany have recognized the plaintiff’s evidentiary problems and started to create rules to facilitate proof. In particular, they created rules of prima facie evidence (Anscheinsbeweis) by which, if public authorities have proven the existence of a cartel, it is assumed that the cartel caused a general price increase and that this price increase had a negative impact on the plaintiff, thus satisfying the second and third requirements for a follow-on suit.[207] Courts also ruled that provisions in a sales contract between a cartel member and the future plaintiff, according to which any antitrust-related damages would be assessed at 15% of the sales price, are enforceable under German contract law, thus addressing the fourth requirement.[208] If a court recognizes these rules of prima facie evidence and enforces contractual fixed-rate damage provisions, it is much easier for plaintiffs to win a follow-on lawsuit.

Interestingly for our purposes, different courts have embraced these procedural twists with different speed and enthusiasm. There is a perception that some courts are more willing to help plaintiffs by shifting burdens of proofs than others.[209] The first court to apply the prima facie evidence rule was the regional court of Dortmund in 2004.[210] Many other regional courts have adopted similar rules since then.[211] One of our interview partners confirmed that differences between courts with regard to evidentiary rules influences plaintiffs’ decisions where to file an antitrust follow-on lawsuit.

The regional court of Mannheim seems to play a special role. This court was consistently named as a plaintiff-friendly venue. One of our interviewees noted a “sensational” (aufsehenerregend) decision in which the court affirmed the enforceability of contractual fixed-rate damage provisions.[212] As one attorney noted:

[In this decision], the court used a double prima facie evidence argument in favor of the plaintiff. . . . This is, of course, a very far reaching evidentiary ruling. If, in addition, you also have this fifteen percent clause in the contract, that’s a safe bet for the plaintiff, isn’t it? And there are many unresolved legal issues where a bright judge, with a creative, solid, well-founded decision, can make a splash.[213]

The interviewee explained that judges in this court seem to care about their reputation in antitrust follow-on proceedings. He told us that, in response to a standard motion to extend a deadline which he had filed in that court, the responsible judge told him:

You know, we are known to schedule proceedings in a speedy manner. We do not want to delay proceedings. We have a reputation to lose.[214]

Antitrust litigation in Mannheim may benefit from spillovers from the court’s prominent role in patent proceedings. In Mannheim, the two chambers that traditionally heard antitrust cases were the same chambers that hear patent cases.[215] The long-time presiding judge of the Mannheim seventh chamber had gained an excellent reputation in patent law, and one interviewee attributed his handling of antitrust follow-on proceedings to his experience in patent law. When the Mannheim regional court created a specialized chamber for antitrust cases in the fall of 2018, this was announced in a press release, in which the minister of justice of the State Baden-Württemberg praised the court as an important economic and legal location factor for the state.[216] The justice ministry also re-appointed a well-known presiding judge from one of Mannheim’s patent chambers to the new antitrust chamber, thereby ensuring that Mannheim’s experience with complex litigation is carried over from patent to antitrust law.

Interviewees deemed other regional courts less plaintiff-friendly than Mannheim. These less-friendly courts include the courts in Kiel, Leipzig and Düsseldorf, as well as the court in Munich. Case management also differs between courts, and Mannheim has a reputation for handling cases more efficiently than other courts.[217]

Given our interviews, we cannot conclude, with the possible exception of Mannheim, that there are clear signs of forum selling in German antitrust follow-on litigation. Even the adoption of the evidentiary rules described above does not necessarily mean that the court adopted them in order to attract litigation. As one interviewee pointed out, given the complexity and volume of antitrust follow-on proceedings, relying on such evidentiary rules expedites proceedings and makes them manageable for the judge. Also, introducing evidentiary rules that help the plaintiff may just be the right thing to do in such cases.

Yet, in December 2018, the German Federal Court of Justice overturned the prima facie evidentiary presumption that lower courts had created, that a cartel causes a general price increase and that this price increase has a negative impact on the plaintiff.[218] Similar to the recent intervention of the Federal Constitutional Court concerning ex-parte injunctions in press law, this could be a sign that the highest German civil court thinks that lower courts have become too plaintiff-friendly in antitrust follow-on lawsuits.[219]

Given our findings in patent law, it may be surprising that we do not find clearer signs of forum selling in German antitrust. After all, these lawsuits are often high-profile cases that are widely reported in the press; they could be a welcome change for a judge who ordinarily has to deal with much smaller cases. Judges could position themselves as guardians of competition vis-à-vis parties, their colleagues, and the wider public. Moreover, the amounts in controversy in these cases are often even higher than in patent cases. As one interviewee put it:

[These are] attractive, very attractive cases, one has to say, compared to the kind of cases a civil chamber has to deal with otherwise. A ten million or even 100 million cartel damage claim is a real tidbit, isn’t it? . . . In addition, [these cases] are just exciting, aren’t they? There are tons of unresolved legal questions, which have to be resolved by the German Federal Court of Justice at some point. And, yes, the facts of the case are mostly exciting. Yes, such a cartel is like a kind of organized crime . . . : secret meeting at the Zurich airport, documents in a safe deposit box, communication through private pre-paid cell phones which do not appear on any corporate account. That’s something, no?[220]

Still, our interviews reveal three reasons why there may be little or no forum selling in antitrust law. First, antitrust follow-on damage lawsuits have only recently emerged in Germany and are still relatively rare. According to estimates by one interviewee, only a few hundred follow-on lawsuits have been filed. Almost no court has to date issued a final ruling on damages; instead, most follow-on court decisions to date have dealt only with liability (Grundurteil).[221] As a result, we simply do not know whether regional courts would differ in the amount of damages they award, and it is difficult to say at this point whether particular German courts are unusually plaintiff-friendly.

Second, the incentives of judges to attract such cases is limited. Antitrust follow-on damage proceedings are typically so complex that it is unclear how a judge could handle them in any effective manner. Judges may be scared off by their sheer size and complexity,[222] especially since the German court system does not adequately reward judges’ work on such complex cases. Judicial performance is typically evaluated, in part, on the number of cases a judge resolves. While the court’s internal case management puts different weights on cases from different areas of the law, the weighing of antitrust follow-on lawsuits may not reflect the actual time required to process such cases. Several of our interviewees reported that a judge dealing with antitrust follow-on cases has between eighteen and twenty-three hours available for such cases, and a judge from the regional court of Dortmund has pinned the number generated by the court’s case management system down to nineteen hours and fifty-three minutes. [223] It is practically impossible to deal with an antitrust case in this amount of time. This time constraint not only seriously lowers incentives for judges to become active in such cases. It also prevents them from developing the expertise required to handle such cases effectively, and lowers incentives for junior judges to move to antitrust chambers.

As a result, some interviewees reported clear signs that judges attempt to avoid antitrust follow-on cases. One interviewee told us of a proceeding where, after nearly nine years of litigation, including a preliminary ruling by to the European Court of Justice, he still does not know whether the court of first instance has jurisdiction or not. One attorney told us:

Given the complexity of cartel damage proceedings, a court is highly motivated at the outset to figure out how to get rid of the case.[224]

Third, court administrators typically do not seem interested in attracting antitrust follow-on litigation to their court. Given the size and complexity of antitrust suits, court fees are unlikely to be high enough to offset the costs.[225] Even more importantly, political considerations may play a role. Antitrust follow-on damage lawsuits are often directed against large German companies. As damage awards against cartel members can amount to hundreds of millions or even to billions of euros, the incentive to encourage such lawsuits are limited. Germany taxes business profits with a trade tax. In 2011, this tax generated revenues of €40.5 billion. This tax generated about 49% of the overall tax income for German municipalities.[226] A large damage award against a German cartel member could significantly lower the trade taxes a German company has to pay. One interviewee noted that this may explain why court administrators do not encourage or facilitate antitrust follow-on lawsuits.

In addition, local sympathies work against follow-on damage lawsuits. A good example may be the pan-European truck manufacturing cartel. One interviewee noted that it may be difficult to convince the ministry of justice in Baden-Württemberg to take a hard stance against the car, truck, sugar, or cement industries, as important manufacturers are located in the state.[227] More generally, German politicians have a long history of protecting the automobile industry,[228] as it generates so many high-skilled jobs. This attitude, which seems widespread in German society, could also have an impact on the attitude of German judges and court administrators.

This analysis shows that while there are some signs of national forum selling in antitrust follow-on damage litigation, these signs are weak, and they are limited to one court, Mannheim.

VI.  International Forum Selling

So far, this Article has focused on competition between German courts. Yet forum selling is not necessarily limited to the national level. It can have an international dimension as well. Also, the weak evidence for forum selling in Germany in antitrust law does not mean that forum selling does not take place. Quite the contrary, we find that forum selling in antitrust is Europe-wide. As such, forum selling occurs on the international, rather than intranational level. It is a competition between the courts of the Netherlands, the United Kingdom, Germany, and other countries. At the same time, pan-European forum selling is less important in the other legal areas investigated by this Article.

International forum selling differs in many ways from national forum selling. While in national forum selling judges (and, in Germany, court administrators) are the most important actors, in international forum selling national legislators play a more prominent role. In addition, both the substantive and procedural rules may vary across courts, and there is often no institution entrusted with ensuring the uniform application of rules in different fora.[229]

A.  Antitrust Law

Antitrust follow-on damage litigation has a strong international component. Cartels often involve member companies from different countries within the European Union. The Court of Justice of the European Union (CJEU) has confirmed that in such cases, a plaintiff may choose one cartel member as an “anchor defendant” and may sue several cartel members at the anchor defendant’s domicile, even if the other companies are domiciled in other E.U. member states.[230] Article 6(3)(b)(2) of the Rome II Convention[231] enables the plaintiff to sue cartel members jointly and severally in the anchor defendant’s domicile not only for damages that occurred in that E.U. member state, but for all damages in the entire European Union.[232] That is, if a company has been harmed by a cartel whose members are domiciled in Germany, the Netherlands, the United Kingdom, and Finland, the plaintiff can choose to sue in any of those four countries. If it chooses Germany and the anchor defendant resides in a city over which the court has jurisdiction, this court can order the German, Dutch, British, and Finnish cartel members to compensate the plaintiff for all the damages the cartel caused to the plaintiff in any of the twenty-eight member states of the European Union. As a result, courts from different member states are in direct competition in antitrust follow-on damage lawsuits.[233]

In this Europe-wide competition, German courts do not fare well.[234] The strongest competitors are the United Kingdom and the Netherlands.[235] Finland is also attractive due to low court and translation costs, highly skilled and motivated judges, and speedy proceedings. While, according to one our interviewees, Germanys offers a highly nuanced and predictable jurisprudence, the German court system has serious disadvantages: limited flexibility as to the language in which documents are submitted,[236] lack of information-technology savviness, and higher costs.[237]

Dutch proceedings are attractive due to their significantly lower costs.[238] According to one interviewee, a Dutch proceeding may cost only €70,000 in court fees for a case that includes two appeals. Dutch judges also engage in more active case management than their German counterparts in order to ensure that cases proceed in a speedy and efficient way.

Courts in the United Kingdom are also an attractive venue for follow-on litigation.[239] American plaintiffs find the United Kingdom particularly hospitable because of the similarity between the American and English legal systems and the availability of discovery.[240] The United Kingdom even created a specialized courtthe Competition Appeal Tribunalto attract such litigation. Despite its name, this court acts as a trial court for antitrust follow-on lawsuits. The Tribunal has not only ensured that judges are experienced in antitrust matters, it has also increased their incentive to attract cases. Following reforms in the Consumer Rights Act of 2015, U.K. law now also features a specialized opt-out class-action mechanism for antitrust law.[241]

Nevertheless, the United Kingdom is not always the ideal venue for antitrust follow-on litigation. The main disadvantage of British proceedings are the immense cost and time of discovery. More generally, proceedings in the United Kingdom are perceived as slow. Furthermore, given the uncertainty generated by Brexit, it is unclear whether U.K. judgments will continue to be enforceable in other E.U. member states. If the United Kingdom leaves the European Union without a treaty requiring the enforcement of U.K. judgments, this could severely impede the ability of U.K. courts to attract pan-European antitrust follow-on litigation.[242]

Comparing the attitudes of court administrators towards antitrust follow-on lawsuits between Germany and the United Kingdom, several interviewees noted that service industries—including legal servicesare an important factor in the United Kingdom both economically and politically. The U.K. government is pushing London as a legal services hub, just as it is pushing the city as a financial hub. According to one interviewee, legislators in the United Kingdom and the Netherlands actively attempt to attract as many antitrust lawsuits as possible. In contrast, German legislators simply do not see litigation as a business opportunity perhaps, as discussed further below, because they fear the impact of antitrust on German manufacturing, which they consider more important than services.[243]

Taken together, our interviews reveal that there are strong signs in antitrust of legislatively-backed forum selling on a European level.

B.  Patent and Press Law

Patent litigation also has a European dimension.[244] Particularly in patent litigation battles that involve global companies and products, litigants can choose among a range of possible venues. Important European countries for patent litigation include Germany, the United Kingdom, the Netherlands, and France.[245]

Nevertheless, the degree to which courts in these countries compete with each other is limited. As described above, antitrust plaintiffs can recover damages that occurred in the entire European Union through a suit in a single European court.[246] Therefore, judgments by courts from different countries can be seen as close substitutes. This is not true for patent law. In most cases, a court can de facto only issue an injunction against the sale of an infringing product in its own jurisdiction. Similarly, it can generally only grant damages for harm that occurred in that jurisdiction.[247] Therefore, a company might need to sue in several jurisdictions in order to fully enforce its right.

The venue decision also turns on considerations that are unrelated to a court’s efforts to attract business, such as the size of the local market. An injunction against distributing a product in a large market is likely to require design and/or manufacturing changes that will, because of economies of scale, affect smaller markets as well, while an injunction in a small country is less likely to have cross border influence. As a result, courts from different European countries are in less direct competition with each other in patent litigation than they are in antitrust follow-on damage litigation.[248]

In accordance with this view, our interviewees generally reported that the German patent system faces limited competitive pressure from neighboring jurisdictions. Suing in Germany is attractive in part because the German market is so large that an injunction against distribution in Germany is likely to terminate the product, to require substantial redesign, or to force the defendant to negotiate a license. Still, interviewees mostly claim the speed and the low cost[249] as well as the highquality of the German legal system are the main reasons that Germany attracts the highest number of patent cases in Europe.[250] Compared to their counterparts in other European countries, German patent courts may also be attractive because they are reluctant to stay infringement proceedings pending the outcome of parallel nullification proceedings. As discussed above, the bifurcated patent system in Germany means that courts in infringement actions do not consider challenges to the validity of a patent, so they can proceed more swiftly.[251]

Outside antitrust and patent law, European forum selling seems to play an even lesser role. In press law, for example, German media publish mostly in German for a German audience. As a result, plaintiffs do not consider courts in other jurisdictions to be a viable alternative to the German courts.

VII.  Generalizing from the Case Studies

While each area of law is different, it is possible to make some generalizations. In some ways forum selling is similar in the United States and Germany, but in other ways it is very different.

A.  Similarities to Forum Selling in the United States

1.  Broad Plaintiff Choice of Forum

The key similarity between the United States and Germany is the importance of permissive venue rules that give plaintiffs the ability to sue in almost any court. Both countries had rules in patent that allowed the plaintiff to sue anywhere a product was sold, which, for most patents of any significance, gave plaintiffs the ability to sue anywhere. In both Germany and the United States, such liberal venue rules led to the clearest examples of forum selling. In the United States, one district, the Eastern District of Texas, actively encouraged cases and was able to garner over a quarter of all patent filings, although a recent Supreme Court decision may end its dominance.[252] In Germany, Düsseldorf and Mannheim have actively competed for patent cases. Venue rules in other areas of potential forum selling, such as press law and antitrust, have also been interpreted to give plaintiffs wide choice of forum.

2.  The Importance of Procedure

In both Germany and the United States, the differences that make some courts more attractive than others are primarily procedural rather than substantive. In both the United States and Germany, key factors in patent litigation are speed and the reluctance of judges to stay proceeding pending decision on patent validity by another administrative or judicial body. Similarly, in press law, speed and the ability to procure ex parte injunctions are crucial factors.

The fact that competition between courts usually relates to procedure rather than substance reflects the less stringent review of procedural decisions. In both the United States and Germany, decision relating to substantive law are generally reviewed rigorously (de novo). In contrast, many procedural decisions are viewed as within the discretion of the judge of first instance and reviewed deferentially. In the United States, most procedural decisions are reviewed for abuse of discretion. In Germany, while not all procedural norms grant a judge discretion, those that do are, reversed only if there is a “mistake in the exercise of discretion” (Ermessensfehler), like such decisions in the United States.[253] In addition, in both the United States and Germany, interlocutory appeals are disfavored, so most procedural issues can only be appealed after the case as a whole has been terminated. Finally, procedural mistakes are unlikely to lead to reversal, as decisions are reversed only if they potentially influenced the outcome of the case (in Germany)[254] or are not “harmless” (in the United States).

3.  Judicial Efforts to Enhance Forum or Even Judge Shopping

In both the United States and Germany, judges seem to have enhanced the ability of plaintiffs to get a hearing before the judges of their choice. Deviations from general principles of random case assignment made it easier for plaintiffs to predict which judge will hear their case, and for judges to position themselves strategically. In the Eastern District of Texas, local rules allowed the plaintiff to sue in the courthouse of its choice, and, since there was often only one judge who heard patent cases in that courthouse, the ability to choose the courthouse meant ability to choose the judge. In Germany, the fact that judges in press law preliminary injunction cases would discuss their likely decision with plaintiff’s counsel and allow the plaintiff to withdraw (and sue elsewhere) if the decision was likely to be adverse gave plaintiffs a rather unique ability to select a favorable forum. More generally, as discussed below, specialization in Germany makes it easier for plaintiffs to choose the judge.

4.  Limits of Forum Selling and the Risk of Backlash

While judges on both sides of the Atlantic may have the means and incentives to make their court attractive to plaintiffs, judges cannot sell their forum without limits. If judges stretch rules too far in their attempt to offer an attractive product to plaintiffs, the legislature or higher courts may intervene to reduce differences between different fora or eliminate venue choice altogether.

 In fact, on both sides of the Atlantic, the highest courts have recently limited the ability or extent to which judges can make their court attractive to plaintiffs: over the last two years, the U.S. Supreme Court has restricted forum choice in patent law cases, the German Federal Constitutional Court has limited ex parte injunctive relief in press law cases, and the German Federal Court of Justice has restricted prima facie evidence rules favoring plaintiffs in antitrust follow-on lawsuits. While these decisions may not stop forum shopping and forum selling altogether, they show that there are limits on the power of lower court judges and administrators to “sell” their court.

B.  Differences from Forum Selling in the United States

1.  The Importance of Judicial Quality

In both the United States and Germany, defenders of courts that succeed in the competition for cases argue that those courts attract cases by offering higher quality adjudications rather than by being biased toward the plaintiff. In the United States, those claims did not bear much scrutiny. For example, if the Eastern District of Texas was attractive on account of the expertise of its judges, it should have been just as attractive to defendants seeking declaratory judgment as to plaintiffs. Yet defendants almost never chose the Eastern District of Texas.[255]

In contrast, quality decisionmaking seems to be an import feature making some German courts more attractive than others. The most popular patent and press courts are respected for the quality of their decisions. Nevertheless, at least for the highest quality courts, there sometimes seems to be a tradeoff between quality and speed: if a party wants a high-quality decision in patent or press law, it may choose a different court than if it wants a fast decision.

Some of the innovations that make particular German courts attractive are also plausibly good rather than simply pro-plaintiff. In antitrust, the evidentiary rules that facilitate follow-on lawsuits seem to be warranted in cases where a public authority has already proven wrongful conduct and where more stringent procedures would mire judges and lawyers in costly and complex proceedings. Similarly, although speed often benefits plaintiffs (particularly in press cases), it is also true that, all other things equal, faster is better than slower.

2.  Specialization Facilitates Forum Selling

Forum shopping and thus forum selling in Germany is made easier by judicial specialization.[256] In the United States, even if one or two judges would like to hear more of a particular kind of case, it is difficult for plaintiffs to be sure their case will be heard by those judges, because most U.S. judges are part of courts with more than a dozen judges, and cases are generally assigned randomly. So, in most circumstances, a plaintiff has no more than a five or ten percent chance of getting a particular judge. In contrast, in Germany, specialized judges generally sit in chambers consisting of only three judges, and all cases of the relevant kind will be assigned to that chamber (or to one of two or three chambers with that specialization).[257] In addition, the dominant influence of the presiding judge ensures that cases assigned to that chambers are likely to be treated similarly no matter which judge actually takes primary responsibility.

Specialization also facilitates forum selling, because it requires coordination among fewer judges. If a U.S. court wants to acquire a reputation as being plaintiff friendly in a particular area, all (or most) of the dozen or more judges on the court must agree. Agreement among such a large group is difficult. In contrast, in Germany, agreement is only necessary among the three judges in the relevant chamber, and agreement is made easier by the outsized influence of the presiding judge.

3.  The Importance of Administration

Court administrators also play a larger role in making some German courts more attractive. The ministry of justice, because of its connection to the government and political parties, has a greater interest in issues such as regional economic development and court revenue than individual judges. When attracting a particular kind of case, such as patent cases, seems likely to benefit the local economy, to increase the state’s reputation, or to bring in fee revenue, the court administration can allocate judges and other resources to that kind of case.

Each court’s executive committee can also play an important role by setting up chambers specializing in areas they want to attract cases. In addition, by reallocating other cases or establishing a second chamber with the same specialization, the executive committee can ensure that judges who are successful in attracting more cases are not overwhelmed by their caseloads. In fact, through its promotion policies, the ministry of justice can ensure that successful judges are rewarded.

4.  Political Economy

Unlike in the United States, there is no separate court system at the federal and the state level in Germany. All courts apply the relevant federal and state law to the cases they hear. Yet, regional courts and courts of appealthe two kinds of courts that are of primary interest in this Articleare established, managed, and financed by each of the sixteen states. German states compete against each other on many dimensions. Because the judicial system is largely managed at the state level, courts can de facto become part of the competitive positioning of a state against other states in the German federation. Conversely, when overly aggressive courts would undermine the competitiveness of local industry or the prospects of state politicians, court administrators at the state level have both the ability and incentive to ensure that courts play a more constructive role.

When considering the political economy of German courts, it is important to note that key decisions are made at the state level, not by individual courts and not at the federal level. For example, the promotion of judges is largely a state-level decision, as is the decision to fund new or improved facilities, or to expand the number of judicial positions in a particular court. While court executive committees have the power to create specialized chambers, this requires the diversion of resources from other legal areas. As a result, the decision to create a specialized press chamber or to add a second patent chamber usually requires the cooperation of state-level administrators.

The prevalence of forum selling seems to vary with the incentives of the court administration and the state. We found the most support for forum selling in patent law, weaker evidence in press law, and even less evidence in antitrust litigation. If one considers the incentives of the court administration and the states in which these courts are located, this outcome is not surprising. If a politician wants to advance her state in competition with other German states, it seems highly promising to promote major cities in her state (such as Düsseldorf in North Rhine-Westphalia) as innovation hubs that are supported by respected patent litigation courts.

It is much less attractive for a politician to push local courts in press or antitrust law. If courts become important litigation hubs in press law, the politician may jeopardize his good relationship with local and national media outlets that may be hurt by the courts. Similarly, if courts become important centers for antitrust litigation, the politician may hurt his relationships with manufacturers, which are a particularly important sector of the German economy.

In other words, in patent law, the plaintiffs who benefit from forum selling are inventors (patent owners), and the defendants who are hurt are alleged patent infringers. In press law, the benefiting plaintiffs are often celebrities, and the hurt defendants are German media outlets. In antitrust law, the benefiting plaintiffs are a large number of companies with relatively small damage claims each, while the hurt defendants are large German employers with large damage liabilities. Politically, it is attractive to help the plaintiffs in patent law, but not in press or antitrust law.

Theoretically, one could imagine a court in a small rural district emulating the U.S. District Court for the Eastern District of Texas and attempting to attract patent or other cases by becoming pro-plaintiff. Based on our interviews, we have found no German court that adopted this strategy. Nor do we deem it likely that any German court would do so for three reasons. First, even a small German regional court is unlikely to be located in areas that are as rural as the Eastern District of Texas. Germany is simply more urban than the United States. There are few sparsely populated areas, and German courts tend to be situated in major cities. Second, the auto industry and other important manufacturers have factories throughout Germany, and much of German media has a national audience. This decreases the likelihood that a judge in a rural regional court would adopt a strategy that could hurt German manufacturing, as it would likely hurt local factories. Similarly, anti-media decisions would likely generate negative press attention that would reach local audiences. Third, and perhaps most importantly, even if an individual judge or court were interested in such a strategy, it would be against the interests of court administrators in the state’s justice ministry, who have the power to decrease funding for the court and to impede the promotion of the judges whose attempts to attract cases might undermine the local economy or the reputation of the courts. Political economy considerations operate strongly at the state level, not solely at the district level. As discussed above, it seems unlikely that a German state would tolerate judicial activity that hurt manufacturing in the state or that resulted in negative press for its politicians.

Conclusion

Judges and court administrators compete for judicial business in Germany, especially in patent and press law cases. They do so for a variety of reasons, including prestige and court revenue. While some ways in which courts compete arguably improve justice by speeding up proceedings and increasing predictability, other ways are more problematic. Failure to stay patent infringement proceedings until validity challenges are resolved by other courts not infrequently subjects patent defendants to unjust infringement judgements based on invalid patents. Similarly, ex parte preliminary injunctions in press cases denies defendants their essential right to be heard. As in the United States, forum selling in Germany is facilitated by loose venue rules that allow most patent, press, and antitrust plaintiffs to sue in any court.

 


[*] *.. Professor of Intellectual Property, Center for Law & Economics, ETH Zurich, Switzerland.

[†] †.. Senior Research Fellow, Max Planck Institute for Research on Collective Goods, Bonn, Germany.

[‡] ‡.. Edward G. Lewis Professor of Law and History, Gould School of Law, University of Southern California, Los Angeles. The authors would like to thank Scott Altman, Konstantin Chatziathanasiou, Luca Enriques, Martin Gelter, Dietmar Harhoff, Stefanie Egidy, Christoph Engel, Michael Gilbert, Michael Halberstam, Orin Kerr, Greg Reilly, John Langbein, Jeff Rachlinski, and David Schwartz, as well as conference participants at the American Law & Economics Conference 2018, the Conference on Empirical Legal Studies 2018, the Economic Analysis of Litigation Conference 2018, the Society for Institutional & Organizational Economics Conference 2018, and seminar participants at Castle Ringberg, the Chicago-Kent Roundtable, ETH Zurich, the Max Planck Institute for Research on Collective Goods (Bonn), Wayne State, New York University, Boston University, and the Universities of Bonn, Glasgow, Oxford and Virginia for helpful feedback. Excellent research, transcription, and translation assistance by Luca Baltensperger, Christine Bircher, Jasmin Büeler, Meike Pauletzki, Helen Reinhart, Daniela Sele, Carlo Sorba, Martin Sternberg, and Miriam Tinner is gratefully acknowledged. In addition, we would like to thank the dozens of attorneys, judges, and court officials we interviewed. Without their insights, this Article could not have been written.

 [1]. See generally Daniel Klerman & Greg Reilly, Forum Selling, 89 S. Cal. L. Rev. 241 (2016) (discussing forum selling in the the United States and early modern England); J. Jonas Anderson, Court Capture, 59 B.C. L. Rev. 1543 (2018) (arguing that the concentration of patent cases in the U.S. District Court for the Eastern District of Texas is a result of capture by special interests); Gerhard Wagner, The Dispute Resolution Market, 62 Buff. L. Rev. 1085 (2014) (discussing interjurisdictional competition, forum shopping, and the merits of bilateral forum selection).

 [2]. See generally Daniel Klerman, Forum Selling and Domain-Name Disputes, 48 Loy. U. Chi. L.J. 561 (2016) (showing that trademark owners select arbitration providers that are more likely to decide cases in their favor in domain name disputes).

 [3]. See generally John Langbein, The German Advantage in Civil Procedure, 52 U. Chi. L. Rev. 823 (1985) (identifying the advantages of the German system of civil procedure).

 [4]. Forum selling may also occur in other areas of German law that have flexible venue rules, including trademark, bankruptcy, unfair competition, criminal and labor law, and in cases related to the Internet. While we also gathered some evidence of forum shopping in some of these areas, this Article will focus on patent, press, and antitrust law.

 [5]. See Frank H. Easterbrook & Daniel R. Fischel, The Economic Structure of Corporate Law 217–18 (1991) (arguing for a race to the top).

 [6]. See Lucian Arye Bebchuk, Federalism and the Corporation: The Desirable Limits on State Competition in Corporate Law, 105 Harv. L. Rev. 1435, 1438 (1992) (“[C]ritics of state competition view it as a ‘race for the bottom,’ in which . . . states are driven to offer rules that benefit managers at the expense of shareholders.” (citation omitted)).

 [7]. Local courts (Amtsgerichte) generally have jurisdiction over civil lawsuits with an amount of controversy less than €5,000. See Gerichtsverfassungsgesetz [GVG] [Courts Constitution Act], May 9, 1975, Bundesgesetzblatt, Teil I [BGBl. I] at 1077, as amended by Act of July 12, 2018 BGBl I at 1151, §§ 23, 71 [hereinafter German Courts Constitution Act]. Cases from local courts are appealed to the regional courts and are subject to a second appeal to the Federal Court of Justice.

 [8]. Id. § 119.

 [9]. Id. § 133.

 [10]. Id. § 75. In practice a chamber might consist of more than three judges, and in each individual case a cast of three judges is selected from all the judges in the chamber.

 [11]. All judges in a chamber, including the presiding judge, can act as reporting judge. Allocation is either determined by pre-established rules of the chamber (under which cases are usually assigned quasi-randomly), or by a discretionary decision of the presiding judge. Section 21g of the German Courts Constitution Act, which requires chambers to establish ex-ante rules on which judges will participate in which proceedings, does not mandate rules predetermining the reporting judge for each case. See Brian Valerius, GVG § 21g [Geschäftsverteilung innerhalb der Spruchkörper], in Beck’scher Online-Kommentar 5 (Jürgen-Peter Graf ed., 28th ed. 2017).

 [12]. For example, the presiding judge determines when to hold an oral hearing, and which witnesses to summon before the court. Zivilprozessordnung [ZPO] [Code of Civil Procedure], Dec. 5, 2005, BGBl. I at 3202, as amended by Act of Jan. 31, 2019, BGBl I, 54, art. 1, §§ 272–73 [hereinafter German Code of Civil Procedure].

 [13]. The Federal Constitutional Court, which publishes dissenting opinions, is a notable exception.

 [14]. The allocation of cases to chambers is usually settled at the beginning of a year for the entire year. See German Courts Constitution Act, supra note 7, § 21e.

 [15]. For appeals involving up to €600, the court of first instance has to expressly allow the appeal under some circumstances. See German Code of Civil Procedure, supra note 12, § 511(2), (4).

 [16]. Id. § 543(1)(1). The regional court of appeals has to allow a decision to be appealed if “the legal matter is of fundamental significance” or “the further development of the law or the interests in ensuring uniform adjudication require a decision” by the Federal Court of Justice. Id. § 543(2).

 [17]. Id. § 544(1); Gesetz, betreffend die Einführung der Zivilprozessordnung [ZPOEG] [Act Concerning the Introduction of the Code of Civil Procedure], Jan. 30, 1877, Reichsgesetzblatt [RGBl] at 244, as amended by Act of June 21, 2018 BGBl I. at 863, § 26(8), https://www.gesetze-im-internet.de/zpoeg/index.html. This effectively means that regional courts and courts of appeals can shield their decisions from review by the Federal Court of Justice by not allowing a decision to be appealed, if no potential subject of a complaint by either party exceeds €20,000.

 [18]. Formally, the ordinary courts of the states are subordinate to the state ministries of justice. In order to guarantee judicial independence, important issues such as the allocation of judges to chambers are not determined by the ministries of justice or the court presidents, but by the court executive committees, which are the self-governing bodies of judges at individual courts.

 [19]. Many of these officials have been judges at one point during their career, but serve at the ministries for at least some amount of time. While they are serving at the ministries, they do not act as judges. This also means that they are expected to follow instructions by their superiors, while of course judges enjoy a relatively high level of independence.

 [20]. Legal education in Germany comprises a university education of normally four to five years and a mandatory legal training (Referendariat) of two years.

 [21]. Deutsches Richtergesetz [DRiG] [German Judiciary Act], Apr. 19, 1972, BGBl. I at 713, as amended by Act of June 8, 2017, BGBl I at 1570, § 48, translation at https://www.gesetze-im-internet.de/englisch_drig/index.html.

 [22]. Rolf Stürner, Rechtspflege durch unabhängige Organe oder Institutionen ein Grundprinzip der Rechtsstaatlichkeit?, 72 Juristen-Zeitung [JZ] 905, 910 (2017).

 [23]. Johann-Friedrich Staats, Richterbeförderung und richterliche Unabhängigkeit in Deutschland: ein systemimmanenter, aber reduzierbarer Konflikt, in Festschrift für Peter Riess zum 70. Geburtstag am 4. Juni 2002 1017, 1020 (Ernst-Walter Hanack et al. eds., 2002) (quoting Prussian minister of justice Adolf Leonhardt). For historical background, see Thomas Ormond, Richterwürde und Regierungstreue: Dienstrecht, politische Betätigung und Disziplinierung der Richter in Preußen, Baden und Hessen 1866–1918, at 507–510 (1994).

 [24]. German Courts Constitution Act, supra note 7, § 21a.

 [25]. Id. § 21e.

 [26]. Id. §§ 1213, 17.

 [27]. See, e.g., German Code of Civil Procedure, supra note 12, § 281.

 [28]. While important decisions by higher German courts are published in official case collections of the courts and are also published widely in practitioner-oriented law journals, less important decisions and, in particular, first instance decisions are usually not published. It is the courts themselves which decide whether to publish a decision or not. In many cases, unpublished court decisions can only be retrieved by filing a request with the respective court to get a physical paper copy of the decision. As a result, there is no comprehensive data on German court decisions available. Given that there is no database in Germany that offers similar coverage to U.S. databases such as Westlaw, LexisNexis, Pacer, or Lex Machina, scholars interested in the German court system have to rely on other methods. In addition to our qualitative interviews, we also performed a literature review in both the legal and the trade press literature.

 [29]. For further discussion on snowball sampling, see Rowland Atkinson & John Flint, Snowball Sampling, in The SAGE Encyclopedia of Social Science Research Methods 1044, 1044 (Michael S. Lewis-Beck et al. eds., 2004).

 [30]. See Yongwook Paik & Feng Zhu, The Impact of Patent Wars on Firm Strategy: Evidence from the Global Smartphone Industry, 27 Org. Sci. 1397, 1412 n.10 (2016) (“Germany has a well-respected specialized patent court system with highly specialized judges for patent infringement cases wherein cases are ruled on relatively quickly, and it can be easier to obtain an injunction in Germany than in other countries. Thus, many technology firms tend to pursue lawsuits in Germany . . . .”). During the so-called “smartphone wars,” Samsung and Apple chose to litigate against each other in Germany. Id. at 1401; Stuart J.H. Graham & Nicolas Van Zeebroeck, Comparing Patent Litigation Across Europe: A First Look, 17 Stan. Tech. L. Rev. 655, 657 (2014).

 [31]. Thomas Kühnen & Rolf Claessen, Die Durchsetzung von Patenten in der EU Standortbestimmung vor Einführung des europäischen Patentgerichts, 115 Gewerblicher Rechtsschutz und Urheberrecht [GRUR] 592, 593 (2013). Katrin Cremers et al., Patent Litigation in Europe, 44 Eur. J.L. & Econ. 1, 23 (2017) [hereinafter Cremers et al., Patent Litigation in Europe] estimate that between 2000 and 2008, Germany attracted four times as many cases as France, the Netherlands, and the United Kingdom combined. If a company has protected its invention not only in Germany, but also in other European countries, the company can choose the country in which it wants to litigate. However, due to the territoriality principle governing patent law, if a company sues in a German court, it can only recover damages that occurred in Germany. As a result, there is no direct competition between patent courts in different European countries at this stage. Düsseldorfer Dominanz, JUVE Rechtsmarkt, Aug. 2018, at 57, 57 estimates that German courts attract 900 to 1000 patent cases per year, while all other European patent courts attract up to 500 cases.

 [32]. See Patentgesetz [PatG] [Patent Act], Dec. 16, 1980, BGBl. I at 1, as amended by Act of Oct. 8, 2017, BGBl. I at 3546, § 143(2), translation at https://www.gesetze-im-internet.de/englisch_ patg/index.html [hereinafter German Patent Act] and the respective laws of the individual states. The twelve designated patent courts are located in Berlin, Braunschweig, Düsseldorf, Erfurt, Frankfurt, Hamburg, Leipzig, Magdeburg, Mannheim, Munich, Nürnberg, and Saarbrücken.

 [33].                German Code of Civil Procedure, supra note 12, § 32.

 [34]. Thomas Kühnen, Patent Litigation Proceedings in Germany: A Handbook for Practitioners 290, 290 (Frank Peterreins trans., 7th ed. 2015) [hereinafter Kühnen, Patent Litigation]; Christian Osterrieth, Patent Enforcement in Germany, in Patent Enforcement Worldwide: Writings in Honour of Dieter Stauder 111, 121–22 (Christopher Heath ed., 3d ed. 2015) [hereinafter Osterrieth, Patent Enforcement in Germany]; Matthias Zigann, Die Sicht des Patentinhabers, in Handbuch des Patentrechts 1033, 1060 (Maximilian Haedicke & Henrik Timmann eds., 2012).

 [35]. TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514, 152021 (2017).

 [36]. Jochen Herr & Marc Grunwald, Speedy Patent Infringement Proceedings in Germany: Pros and Cons of the Go-To Courts, 7 J. Intell. Prop. L. & Prac. 44, 44–45 (2012); Fabian Gaessler & Yassine Lefouili, What to Buy When Forum Shopping? Analyzing Court Selection in Patent Litigation 12, (Toulouse Sch. of Econ. Working Papers, Paper No. 17775, 2017), http://ssrn.com/abstract=29313

51.

 [37]. See supra Section I.A.

 [38]. Several scholars have estimated the allocation of patent cases among the regional courts. Gaessler and Lefouili show that, during the period from 2003 to 2008, Düsseldorf heard almost ten times as many cases as Munich. Gaessler & Lefouili, supra note 36, at 19; see also Katrin Cremers & Paula Schliessler, Patent Litigation Settlement in Germany: Why Parties Settle During Trial, 40 Eur. J.L. & Econ. 185, 197 (2015); Mathieu Klos, Standortvorteil, JUVE Rechtsmarkt, Apr. 2010, at 72, 72–73 [hereinafter Klos, Standortvorteil]; Kühnen & Claessen, supra note 31, at 593. Stephan Hase, Die statistische Erfassung von Rechtsstreitigkeiten in Patent-, Gebrauchsmuster- und Arbeitnehmererfinder-Sachen in der BRD im Jahre 1992 – der Abschluß einer Versuchsreihe, 84 Mitteilungen der deutschen Patentanwälte [MDP] 329, 330 (1994) reports that the regional court of Düsseldorf dealt with 201 cases in 1992, while the Mannheim court dealt with 52 and the Hamburg court dealt with 25. No data on the Munich court was available. Osterrieth, supra note 34, at 115–16, estimates that of the more than 1300 patent cases that are filed in Germany per year, about 500 cases are filed in Düsseldorf, about 280 in Mannheim, and about 170 each in Munich and Hamburg, although he does not provide data to substantiate his claim. The German Federal Statistics Office’s data indicates how many patent cases the four busiest regional courts, as well as the Hamburg court, handled over the last years:

Note: Statistisches Bundesamt, Rechtspflege: Zivilgerichte 2017, Fachserie 10, Reihe 2.1 (2018), https://www.destatis.de/GPStatistik/receive/DEHeft_heft_00083616. This data should be interpreted with care. They capture all lawsuits regional courts handled in a given year that primarily concerned patents, utility models, and employee inventions. They include all cases that were concluded in a year, either because the court decided the case, or because the parties settled the case, or for other reasons. Data before 2010 is not reliable because of changes in the data coding structure. Nevertheless, this data gives an indication of the relative market share of German patent courts over the last few years. Recently, the Düsseldorf district court started to publish case numbers itself, with other courts adopting the same practice immediately. In 2017, Düsseldorf heard 488 cases, Mannheim 215, and Munich 181. Düsseldorfer Dominanz, supra note 31, at 57.

 [39]. Düsseldorfer Dominanz, supra note 31, at 57.

 [40]. See also Gaessler & Lefouili, supra note 36, at 12.

 [41]. For example, the distribution of case numbers could be the result of network externalities: potentially, a court with high numbers of past cases in the past is more attractive to future plaintiffs than a court with no experience at all in a certain field of law. On the role of network externalities in the law, see generally Mark A. Lemley & David McGowan, Legal Implications of Network Economic Effects, 86 Calif. L. Rev. 479 (1998).

 [42]. Klerman & Reilly, supra note 1, at 241.

 [43]. Cremers et al., Patent Litigation in Europe, supra note 31, at 13, 27; Herr & Grunwald, supra note 36, at 44; see also Kühnen & Claessen, supra note 31, at 593.

 [44]. Cremers et al., Patent Litigation in Europe, supra note 31, at 14; see also Kühnen & Claessen, supra note 31, at 593.

 [45]. Under German patent law, users of an invention might have a right of prior use, German Patent Act, supra note 32, § 12, or the right to ask for a compulsory license, id. § 24. Also, courts have held that antitrust law in some cases requires a patentee to grant a compulsory license to alleged infringers. See, e.g., Bundesgerichtshof [BGH] [Federal Court of Justice] May 6, 2009, 111 GRUR 694, 2009.

 [46]. Thomas Kühnen, Handbuch der Patentverletzung 446–49 (10th ed. 2018) [hereinafter Kühnen, Handbuch der Patentverletzung]; Peter Mes, Patentgesetz, Gebrauchsmustergesetz § 139 n.42 (4th ed. 2015). One commentator traces this back to the distinction between common law and equity, which does not exist in civil law countries. Ansgar Ohly, “Patenttrolle” oder: Der patentrechtliche Unterlassungsanspruch unter Verhältnismäßigkeitsvorbehalt? Aktuelle Entwicklungen im US-Patentrecht und ihre Bedeutung für das deutsche und europäische Patentsystem, 57 Gewerblicher Rechtsschutz und Urheberrecht: Internationaler Teil [GRUR Int.] 787, 793 (2008). In recent years, a number of commentators have argued that injunctive relief under German patent law should be constrained in some cases. Thomas F. Cotter, Comparative Patent Remedies: A Legal and Economic Analysis 243–48 (2013); Clemens-August Heusch, Der patentrechtliche Unterlassungsanspruch, in Festschrift für Wolfgang von Meibom zum 65. Geburtstag 135, 147 (Christian Harmsen et al. eds., 2010); Marcus Sonnenberg, Die Einschränkbarkeit des patentrechtlichen Unterlassungsanspruchs im Einzelfall 173–236 (2014); Torsten Körber, Machtmissbrauch durch Erhebung patentrechtlicher Unterlassungsklagen? Eine Analyse unter besonderer Berücksichtigung standardessentieller Patente, 59 Wettbewerb in Recht und Praxis [WRP] 734, 740 (2013); Ohly, supra at 795–97; Christian Osterrieth, Patent-Trolls in Europa: Braucht das Patentrecht neue Grenzen?, 111 GRUR 540, 542 (2009); Hans-Jürgen Papier, Verfassungsrechtliche Anforderungen an den Patentschutz, 8 Zeitschrift für Geistiges Eigentum [ZGE] 431, 432 (2016).

 [47]. See also Cremers et al., Patent Litigation in Europe, supra note 31, at 34–35; Herr & Grunwald, supra note 36, at 44–45. Recently, the ease with which patentees can obtain an injunction became the subject of severe criticism by the former Chief Justice of the German Constitutional Court, who challenged this feature as unconstitutional. Papier, supra note 46, at 432.

 [48]. German Patent Act, supra note 32, §§ 21, 59, 81; see also Graham & Van Zeebroeck, supra note 30, at 67072.

 [49]. German Patent Act, supra note 32, § 143(1).

 [50]. Anja Schwarz, Nullity Proceedings, in Patent Litigation in Germany 227, 228 (Claudia Milbradt ed., 2012); Katrin Cremers et al., Invalid but Infringed? An Analysis of the Bifurcated Patent Litigation System, 131 J. Econ. Behav. & Org. 218, 220 (2016) [hereinafter Cremers et al., Invalid but Infringed?]; Sven Hetmank, The Principle of Separation in German Patent Law and its Implications for Patent Litigation, 34 Eur. Intell. Prop. Rev. 369, 369 (2012).

 [51]. German Code of Civil Procedure, supra note 12, § 148. See infra Section III.B.3 for a description of the practice of the courts in this regard.

 [52]. Cremers et al., Invalid but Infringed?, supra note 50, at 221; Cremers et al., Patent Litigation in Europe, supra note 31, at 13, 27.

 [53]. Cremers et al., Invalid but Infringed?, supra note 50, at 220; Uwe Scharen, The Practice of Claiming Injunctive Relief for Patent Infringement in Germany, 14 J. Intell. Prop. L. & Prac. 112, 116 (2018).

 [54]. In principle, it is possible for a defendant to file for (negative) declaratory relief. However, plaintiffs can counter such maneuver by suing for infringement elsewhere, which would render the defendant’s case inadmissible. This might be different if an action for negative declaratory relief is filed with the courts in another country. See Cotter, supra note 46, at 253–55; Mario Franzosi, Worldwide Patent Litigation and the Italian Torpedo, 7 Eur. Intell. Prop. Rev. 382, 382 (1997); Michele Giannino, Italian Torpedo Actions Can Sink Cross-Border Patent Infringement Proceedings, 9 J. Intell. Prop. L. & Prac. 172, 173 (2014).

 [55]. See also Gaessler & Lefouili, supra note 36, at 31; Herr & Grunwald, supra note 36, at 47.

 [56]. Mathieu Klos, Angriff auf die Bastion, JUVE Rechtsmarkt, June 2012, 83, 87.

 [57]. Osterrieth, Patent Enforcement in Germany, supra note 34, at 115, reports that it takes the regional court of Mannheim about six to eight months after filing a lawsuit to schedule an oral hearing on the merits, while it takes the regional court of Düsseldorf between fourteen and eighteen months. See also Thorsten Bausch & Esther Pfaff, Das “neue Münchner Verfahren” eine Trumpfkarte für den Gerichtsstandort München?, 103 MDP 97, 98 (2012).

 [58]. Klos, Standortvorteil, supra note 38, at 76.

 [59]. See also Gaessler & Lefouili, supra note 36, at 22. Note that these preferences do not seem to be strongly influenced by where companies in these industries are located.

 [60]. Reportedly, courts vary with regard to how quickly they require a plaintiff to file for a preliminary injunction. The courts in Düsseldorf and Hamburg have a reputation for being rather accommodating in this regard, while Mannheim considers late requests for preliminary injunctions inadmissible.

 [61]. Düsseldorf until recently required all documents, including exhibits, to be translated into German. In Mannheim, only briefs have to be filed in German.

 [62]. Note that, in Germany, there are no formal rules of precedent. Therefore, there is no difference between published and unpublished decisions regarding whether they are considered binding precedent.

 [63]. This might have changed in recent years.

 [64]. Mannheim normally does not schedule an early hearing but aims at resolving cases after only one oral hearing following the exchange of briefs between the parties. In Düsseldorf, traditionally an oral hearing was scheduled early during the proceedings in order to deliberate the procedural schedule, with a second oral hearing taking place after the end of the written procedure. Reportedly, judges have recently adopted Mannheim’s approach in at least some proceedings. Munich, like Düsseldorf, schedules early hearings but aims to provide parties with a preliminary assessment of the merits of a case. See Osterrieth, Patent Enforcement in Germany, supra note 34, at 114–15; Herr & Grunwald, supra note 36, at 46. See also Hinweise zum Münchner Verfahren in Patentstreitsachen, Regional Court of Munich, (Dec. 2016), https://www.justiz.bayern.de/media/images/behoerden-undgerichte/infoblatt_m_nchner_verfahr en__stand_12_2016_.pdf.

 [65]. A “chamber” is a group of judges, usually three, who handle a particular kind of case. So, by increasing the number of patent chambers, court administrators are increasing the number of positions for patent judges.

 [66]. See Gaessler & Lefouili, supra note 36, at 19–22, which shows that, for the time period from 2003 to 2008, average durations of proceedings were considerably shorter in Mannheim as compared to Düsseldorf. Note, however, that the subsequent analysis in Gaessler & Lefouili challenges this notion.

 [67]. See Klerman & Reilly, supra note 1, at 265.

 [68]. If they do not have the technical understanding necessary to establish the facts of the case, German judges are, in fact, under an obligation to appoint an expert. Although judges in principle enjoy large discretion whether to appoint an expert witness, it may give rise to a successful (second) appeal if they do not appoint an expert opinion despite not having the required technical expertise to properly establish the facts of a case. See Kühnen, Patent Litigation, supra note 34, at 725–26; Walter Zimmermann, in 2 Münchener Kommentar zur Zivilprozessordnung § 402, ¶ 7 (Thomas Rauscher & Wolfgang Krüger eds., 5th ed. 2016).

 [69]. Note that expert witnesses are almost always appointed by the court under German civil procedure law. Party-appointed experts, by contrast, play a much less important role than in U.S. civil procedure. See Langbein, supra note 3, at 835.

 [70]. Kühnen, Patent Litigation, supra note 34, at 726.

 [71]. Kühnen & Claessen, supra note 31, at 595; see also Herr & Grunwald, supra note 36, at 46.

 [72]. See Kühnen & Claessen, supra note 31, at 597.

 [73]. Herr & Grunwald, supra note 36, at 46; Marcus Creutz, Richter machen kurzen Prozess, Handelsblatt, Jan. 27, 2011, at 55, 55. The findings in Gaessler and Lefouili largely confirm the statements of our interviewees. Gaessler & Lefouili, supra note 36, at 25. They find that during the time period from 2003 to 2008, the regional court in Düsseldorf relied on expert witnesses least often, followed by Mannheim, while in Munich expert witnesses were involved most often. Id.

 [74].  Interview with attorney (July 8, 2014).

 [75]. Hinweise zum Münchner Verfahren in Patentstreitsachen, supra note 64; see also Bausch & Pfaff, supra note 57, at 97; Herr & Grunwald, supra note 36, at 46; Klos, Standortvorteil, supra note 38, at 75.

 [76]. Klos, Standortvorteil, supra note 38, at 73.

 [77].  Interview with judge (Aug. 14, 2014).

 [78].                Interview with judge (July 9, 2014).

 [79]. See supra text accompanying notes 4850.

 [80]. See supra Section III.A.

 [81]. See Gaessler & Lefouili, supra note 36, at 14.

 [82]. See supra Section III.A.

 [83]. Bundesgerichtshof [BGH] [Federal Court of Justice] Sept. 16, 2014, 116 GRUR 1237, 1238 (2014); Bundesgerichtshof [BGH] [Federal Court of Justice] Nov. 11, 1986, 89 GRUR 284, 284 (1986) (“some chance of success”); Klaus Grabinski & Carsten Zülch, in Patentgesetz: Kommentar § 139, 107 (Georg Benkard ed., 11th ed. 2015).

 [84]. An anecdote may exemplify the attitude: according to this anecdote, a law clerk at a German patent court prepared a meeting with his judge concerning a patent infringement lawsuit. When the clerk turned to the files concerning the invalidity proceedings, his judge told him: “You don’t have to read them, we will not stay proceedings anyhow.” Malte Köllner et al., 95 Thesen zur Aussetzung, 109 MDP 8, 11 (2018).

 [85]. Kühnen & Claessen, supra note 31, at 594; see also Kühnen, Patent Litigation, supra note 34, at 519; Thomas Kühnen, The Bifurcation System in German Practice, in 16th Symposium of European Patent Judges 59, 67 (Eur. Patent Office ed., 2013); Cremers et al., Invalid but Infringed?, supra note 50, at 221; Herr & Grunwald, supra note 36, at 44 (estimating that around 10 to 15 percent of infringement cases are stayed); Scharen, supra note 53, at 120 (noting that a review of the relevant case suggests that motions to stay infringement proceedings were rejected “in the vast majority of cases”); Tobias Wuttke & Peter Guntz, Wie weit reicht die Privilegierung des Klägers durch das Trennungsprinzip?, 103 MDP 477, 483 (2012) (estimating that proceedings were stayed in less than ten percent of cases). In an attempt to verify such claims, in November 2017 we obtained all decisions by one of the three major German patent courts which mention § 148 of the German Civil Procedure Code from the commercial IP database Darts-IP (www.darts-ip.com). This search yielded a total of 737 results, with 536 decisions by the regional court in Düsseldorf, 61 decisions by the regional court in Munich, and 140 decisions by the regional court in Mannheim. Four hundred thirty-four of these decisions were issued between 2010 and 2017. In this time period, the Düsseldorf court decided to stay a proceeding in only 1 out of 267 cases (<1%), the Mannheim court decided to stay a proceeding in 33 out of 124 cases (27%), and the Munich court decided to stay a proceeding in 7 out of 48 cases (15%). Interestingly, during a slightly earlier period (2000 and 2009), the rate of stayed proceedings in Munich was somewhat higher, with 2 out of 10 cases stayed (20%). This might indicate a shift at the time of the introduction of the New Munich Proceedings in 2009. It is important to note, however, that these numbers are unreliable because only a small number of cases were included in the Darts-IP database before 2010 and because some of the stays may have been in related proceedings (that is, of the 33 decisions to stay a proceeding issued by the regional court of Mannheim in the period from 2010 to 2017, 17 were issued in 2013, pointing to the possibility that these cases may have been related).

 [86]. Ulrike Voß, in Prozesskommentar zum Gewerblichen Rechtsschutz § 940 ZPO 1484, 148489, ¶¶ 121–37 (Philipp Moritz Cepl & Ulrike Voß eds., 2d ed. 2018); Herr & Grunwald, supra note 36, at 46 (noting that the regional court of Düsseldorf “rarely stays infringement proceedings, while in Mannheim and Munich a stay of the proceedings, while clearly the exception, may happen slightly more often”); Scharen, supra note 53, at 119.

 [87]. Hinweise zum Münchner Verfahren in Patentstreitsachen, supra note 64. The courts differ somewhat in the test they purport to apply in their decision whether to stay a proceeding. In the decisions we obtained from Darts-IP, see discussion supra note 85, Düsseldorf and Munich between 2010 and 2017 mostly applied the standard set by the Federal Court of Justice in 2014. See Bundesgerichtshof [BGH] [Federal Court of Justice] Sept. 16, 2014, 116 GRUR 1237 (2014) (setting this standard). Since 2015, the Düsseldorf court shifted to demanding that the defendant show that an invalidation is sufficiently likely (hinreichend wahrscheinlich). The regional court in Mannheim mostly asked for a high probability (hohe Wahrscheinlichkeit) of an invalidation. The fact that Mannheim as the regional court which professed to apply the strictest standard also stayed the highest number of proceedings puts into question whether the wording of the test is of any relevance. See also Voß, supra note 86, at 148889, ¶ 137.

 [88]. Creutz, supra note 73, at 55. The Federal Court of Justice has never reverted a decision not to stay a proceeding by the lower courts. One potential reason for this apparent contradiction is the fact that decisions by the lower courts are largely shielded from appellate review, see infra text accompanying notes 98100.

 [89]. Cremers et al., Invalid but Infringed?, supra note 50.

 [90]. Id. at 234.

 [91]. Kühnen & Claessen, supra note 31, at 594. It is important to note that the available data do not provide conclusive evidence on whether German patent courts should be more willing to stay proceedings. In an influential article, the presiding judge of the patent senate at the Federal Court of Justice (Bundesgerichtshof) argued that German patent courts do not stay proceedings at lower than the optimal rate. Peter Meier-Beck, Überlegungen zum Übereinkommen über ein Einheitliches Patentgericht und zur Zukunft des Trennungsprinzips in Deutschland, 117 GRUR 929, 931 n.3 (2015). Unfortunately, Judge Meier-Beck’s analysis is subject to various methodological problems. The dataset he uses does not include information on how often requests to stay proceedings are actually filed and what percentage of these requests are actually granted by German patent courts. As a result, Judge Meier-Beck relies on a number of problematic assumptions in order to calculate the rate at which requests to stay proceedings are granted. The same is true for the optimal rate. It is reasonable to assume that the optimal rate to stay proceedings concerning particular patents should be roughly equivalent to the hypothetical rate at which these patents would be invalidated by the European Patent Office, the German Patent & Trademark Office or the German Federal Patent Court. Unfortunately, data on this hypothetical rate is not available, and estimating such rate is challenging. The main reason for this is that many cases settle between the decision by the regional court on whether to stay the case and the decision by the patent offices or the Federal Patent Court on the validity of the patent. Cremers et al., Invalid but Infringed?, supra note 50. Note that, even using the best dataset that is available on German patent litigation (the Max Planck patent litigation dataset, covering all patent infringement cases at the courts of Düsseldorf, Mannheim and Munich between 2000 and 2008), we were unable to answer this question on the basis of a methodologically sound quantitative analysis.

 [92]. Gaessler and Lefouili, supra note 36, at 21, report that for the time period between 2003 and 2008 proceedings in Munich were stayed in 30% of all cases in which a parallel validity proceeding was pending. By contrast, the regional courts in Mannheim and Düsseldorf only stayed 18% of all cases with parallel invalidity proceedings. Id.; see also Creutz, supra note 73, at 55.

 [93]. Interview with judge (Aug. 14, 2014).

 [94]. The other changes included a standardized procedural schedule aimed at providing plaintiffs with a good estimate for how long it would take them to obtain a judgment. Klos, Standortvorteil, supra note 38, at 73.

 [95]. Interview with attorney (July 4, 2014).

 [96]. Id. (internal quotation marks omitted).

 [97]. Klerman & Reilly, supra note 1, at 250, 278, 301–02.

 [98]. Grabinski & Zülch, supra note 83, § 139, ¶ 108; Karl Harraeus, Über den gleichzeitigen Ablauf von Patentverletzungs- und Patentnichtigkeitsverfahren, 66 GRUR 181, 182 (1964); Thomas Kaess, Die Schutzfähigkeit technischer Schutzrechte im Verletzungsverfahren, 111 GRUR 276, 277 (2009).

 [99]. Grabinski & Zülch, supra note 83, § 139, ¶ 108; Scharen, supra note 53, at 120.

 [100]. See Falk Freiherr von Maltzahn, Die Aussetzung im Patentverletzungsprozeß nach § 148 ZPO bei erhobener Patentnichtigkeitsklage, 87 GRUR 163, 165 (1985).

 [101]. Reto M. Hilty & Matthias Lamping, Trennungsprinzip Quo vadis, Germania, in 50 Jahre Bundespatentgericht: Festschrift zum 50-jährigen Bestehen des Bundespatentgerichts am 1. Juli 2011, 255, 260–262 (Achim Bender et al. eds., 2011); Kühnen & Claessen, supra note 31, at 595; see also Kühnen, Handbuch der Patentverletzung, supra note 46, at 702; Papier, supra note 46, at 443. Arguments against the restrictive interpretation of the requirements for staying proceedings are not new. See Harraeus, supra note 98, at 182–85 (reviewing arguments in favor and against a restrictive interpretation of the requirements for staying proceedings).

 [102]. See Tobias J. Hessel & Maximilian Schellhorn, Die Rückabwicklung des vorläufig vollstreckten Unterlassungstitels im Patentrecht, 119 GRUR 672, 674 (2017); Rüdiger Rogge, Zur Aussetzung in Patentverletzungsprozessen, 45 GRUR Int. 386, 386 (1996).

 [103]. This is the case if the decision by the regional court is confirmed by the regional court of appeals, or if the defendant does not appeal the decision by the regional court. Rogge, supra note 102, at 389; Zigann, supra note 34, § 11, ¶ 537.

 [104]. As mentioned above, Cremers et al., Invalid but Infringed?, supra note 50, at 234, find that a substantial number of challenges to the validity of patents are withdrawn after an infringement decision.

 [105]. See generally Hilty & Lamping, supra note 101 (arguing that the bifurcation principle unfairly favors the plaintiff, and arguing in favor of its abolition); Papier, supra note 46 (arguing that the bifurcation principle has contributed to a situation in which defendants’ constitutional rights are violated). Note that Papier, in his criticism of the German patent litigation system, provides a different reason for the handling of requests to stay proceedings which does not take into account forum selling: “[t]he reason likely is that the civil courts, in an attempt to protect the efficiency of the patent as an exclusive right and a right of defense, are not willing to condone infringing actions during the—often longer—time a nullification proceeding is pending.” Papier, supra note 46, at 440.

 [106]. Klerman & Reilly, supra note 1, at 263–65.

 [107]. See generally Klerman & Reilly, supra note 1 (examining forum selling the United States and proposing solutions to reduce the phenomenon).

 [108]. Note that some of our interviewees indicated that not all judges might want to work exclusively in patent law.

 [109]. Interview with attorney (July 1, 2014).

 [110]. Klerman & Reilly, supra note 1, at 254–57.

 [111]. See generally Nuno Garoupa & Tom Ginsburg, Judicial Reputation: A Comparative Theory (2015) (discussing the importance of reputation for judicial systems in general). In December 2018, for example, the Munich district court banned Apple from selling iPhones 7 and 8 in Germany as long as Qualcomm posted a court-set bond of €1.34 billion (an estimate of Apple’s damages should the appellate court overturn the ban). Qualcomm posted the billion-dollar bond within two weeks, and the judge and his court decision made international news. Martin Coulter & Tobias Buck, Apple Faces iPhone Ban in Germany Over Patent Case, Fin. Times (Dec. 20, 2018), https://www.ft.com/content/c66e9a96-0469-11e9-9d01-cd4d49afbbe3.

 [112]. Interview with attorney (July 1, 2014).

 [113]. Interview with attorney (July 1, 2014).

 [114]. While such a Stuhlurteil is legal, it is extremely rare in complex disputes. Usually, judgments are issued in writing several weeks or months after the oral hearing.

 [115]. Klos, Standortvorteil, supra note 38, at 79.

 [116]. Langbein, supra note 3, at 851.

 [117]. The Unified Patent Court is a proposed common court of EU member states which, once established, will have jurisdiction to hear proceedings regarding the infringement and the validity of European-wide patents. The international agreement by which the Unified Patent Court is to be established was signed in February 2013. Agreement on a Unified Patent Court, 2013 O.J. (C 175) 1. The agreement, however, has not yet entered into force because it has not been ratified by Germany, among other countries.

 [118]. Mathieu Klos, Europäisches Richterranking: Favoriten für Paris, JUVE Rechtsmarkt, June 2017, at 72, 81 points out that becoming a judge in the Unified Patent Court is not only attractive because one may shape both a new court and European patent law. It can also be attractive for financial reasons: while the president of a chamber at a regional court in Germany has a base salary of about €88,000 per year, a Unified Patent Court judge is expected to earn up to €144,000 per year. This is a higher salary than the base salary of a judge at the German Federal Court of Justice.

 [119]. Stürner, supra note 22, at 909.

 [120]. Some prominent patent judges, such as Thomas Kühnen from the court of appeals in Düsseldorf, publish treatises and law journal articles on patent law. Revenue from such activities can be substantial. A recent German Parliamentary inquiry revealed that over 30 judges at the Federal Court of Justice (Bundesgerichtshof) have generated extra incomes between €10,000 and €240,000 per year by publishing legal treatises. See Antwort der Bundesregierung, Deutscher Bundestag: Drucksachen [BT] 18/10781 10, 16, http://dipbt.bundestag.de/doc/btd/18/107/1810781.pdf.

 [121]. Interview with judge (July 9, 2014).

 [122]. See infra Section III.B.1; Klaus Schülke, Patentgerichtsbarkeit, in Die Praxis des Richterberufs 67, 77 (Peter-Christian Müller-Graff & Herbert Roth eds., 2000); Bausch & Pfaff, supra note 57, at 97. In fact, the staffing of patent chambers has become a matter of state politics in the state of North Rhine-Westphalia, where Düsseldorf is located. After the state election in 2005, the ruling political parties (the Christian Democratic Union and the Free Democratic Party) declared in their coalition agreement that the staffing of the two patent chambers at the regional court of Düsseldorf should be increased. Reportedly, the political goal was to secure court revenues and to head off increasing competition from the regional court of Mannheim. See Klos, Standortvorteil, supra note 38, at 83.

 [123]. Interview with judge (July 9, 2014).

 [124].                Court fees in Germany can be substantial, with individual proceedings in a court of first instance generating up to €329,208 in fees. Gerichtskostengesetz [GKG] [Act on Court Fees], Feb. 27, 2014, BGBl. I at 154, as amended by Act of July 12, 2018 BGBl I at 1151, §§ 3(2). 34(1), 39(2), Anlage 1, [hereinafter German Act on Court Fees], http://www.gesetze-im-internet.de/gkg_2004/index.html.

 [125]. Interview with judge (July 16, 2014).

 [126]. Klos, Standortvorteil, supra note 38, at 83. The Düsseldorf estimate is only a rough one. It is based on the observation that the court heard 560 patent cases in 2009, that the typical amount in controversy in German patent disputes is about one million euros, and that the court fees in such a case are about €13,400 if the court issues a final decision. The estimate does not take into account settlements, which may lower court fees.

 [127]. Marcus Jung & Jörn Poppelbaum, Premiumlagen, JUVE Rechtsmarkt, June 2015, at 36, 37.

 [128]. Schülke, supra note 122, at 77.

 [129]. Examples that were mentioned in our interviews include law firms creating jobs in the city.

 [130]. One of the reasons may be that business methods have never been patentable under German or European patent law.

 [131]. Einheitliches Europäisches Patentgericht: Einrichtung einer Lokalkammer am Standort Mannheim nimmt Fahrt auf, Ministerium der Justiz und für Europa Baden-Württemberg (Mar. 1, 2016), http://www.jum.baden-wuerttemberg.de/pb/,Lde/Startseite/Service/Einheitliches+Europaeisch es+Patentgericht_+Einrichtung+einer+Lokalkammer+am+Standort+Mannheim+nimmt+Fahrt+auf/?LISTPAGE=1825757.

 [132]. NRW-Landeshauptstadt auf dem Weg zum Weltpatentgericht Bundesrat berät über Europäisches Patentgericht in Düsseldorf, Staatskanzlei des Landes Nordrhein-Westfalen (Feb. 10, 2017), https://www.land.nrw/de/pressemitteilung/nrw-landeshauptstadt-auf-dem-weg-zum-weltpat entgericht-bundesrat-beraet-ueber.

   [133].               Id.

 [134]. Jung & Poppelbaum, supra note 127, at 37.

 [135]. Id. at 38.

 [136]. Von Schiedsgerichten können wir viel lernen, JUVE Rechtsmarkt, Mar. 2018, at 61, 63 (reproducing an interview with Anne-José Paulsen, the former president of the Düsseldorf Higher Regional Court). Responding to the assertion that “[c]oncerning international orientation, a German courts of appeals cannot compete with large law firms,” she replied:

Yes, we certainly can! Take our patent law chambers at the district and appellate level. Global disputes and the contact with international companies are common practice. This also applies to judges at chambers who deal with antitrust damage lawsuits. It is also common practice that judges travel to international conferences. Recently, several judges—including myself—traveled to China in order to visit IP courts in several large cities. We have many international connections.

 Id.

 [137]. See supra note 117 and accompanying text.

 [138]. The planned Unified Patent Court will consist of different courts. Local branches of its court of first instance are to be set up in various European Union member states. Because of the volume of patent litigation in German courts, Germany was given the right to set up four local branches, with the exact location of the local branches determined by the German government. See Agreement on a Unified Patent Court, art. 7, § 4, 2013 O.J. (C 175) 1, 4. In 2014, the federal minister of justice announced that local branches would be set up in Düsseldorf, Mannheim, Munich, and Hamburg. Standorte für künftige Lokalkammern stehen fest, Legal Trib. Online (Mar. 19, 2014), https://www.lto.de/recht
/nachrichten/n/eu-patentgericht-lokalkammer-standorte-hamburg-duesseldorf-mannheim-muenchen.

 [139]. One rather extreme example of such behavior is a series of decisions in which this judge accused lawyers of lying to the court about the true value of a dispute in an attempt to keep court fees low. Oberlandesgericht Düsseldorf [OLG Düsseldorf] [Court of Appeals Düsseldorf], Apr. 4, 2010, 10 Neue Juristische Online-Zeitschrift [NJOZ] 2425, 2426; OLG Düsseldorf [Court of Appeals Düsseldorf], May 10, 2011, 11 GRUR-Rechtsprechungsreport 341, 342; see also Jürgen Wessing & Eren Basar, Streitwertangabe: strafbar?, 114 GRUR 1215, 1215 (2012); Pia Lorenz, Ein Berufsstand unter Generalverdacht, Legal Trib. Online (Aug. 11, 2011), https://www.lto.de/recht/job-karriere
/j/olg-vorwuerfe-an-anwaelte-ein-berufsstand-unter-generalverdacht.

 [140]. Besides individuals, corporations can also under some circumstances sue media outlets over their reporting, for example in case they spread false information about them. Nevertheless, most cases are brought by individuals.

 [141]. Defendants in practice have no ability to influence the court in which the case is heard.

 [142]. In addition, in some cases, individuals ask the court to order the defendant to publish a reply or correction or sue for damages, or both. Press law in Germany, in principle, follows general principles developed in tort law (Deliktsrecht). The right to publish a reply is an exception and forms part of the laws of the individual states. See, e.g., Hamburgisches Pressegesetz [Press Law of Hamburg], Jan. 29, 1965, Hamburgisches Gesetz- und Verordnungsblatt [Hmb GVBl] at 15, as amended by Act of Oct. 16, 2007, Hmb GVBl at 385, § 11; Landespressegesetz NRW [State Press Law of North Rhine-Westphalia], May 24, 1966, Gesetz- und Verordnungsblatt Nordrhein-Westfalen [GV NRW] at 340, last amended by Act of May 8, 2018, GV NRW at 214, § 11.

 [143]. In contrast, lawsuits seeking to enforce an individual’s right of reply generally have to be brought at the place of business of the relevant publishing house.

 [144]. Disputes in press law are usually heard by regional courts and not local courts because the amount in controversy exceeds €5,000. See German Courts Constitution Act, supra note 7, §§ 23, 71.

 [145]. The German Code of Civil Procedure allows the plaintiff to bring a case before “the court in the district in which the tort was committed.” German Code of Civil Procedure, supra note 12, § 32.

 [146]. Uwe Jürgens, Abgestürzte Gerichtsstände – Der fliegende Gerichtsstand im Presserecht, 67 Neue Juristische Wochenschrift [NJW] 3061, 3064 (2014) [hereinafter Jürgens, Abgestürzte Gerichtsstände].

 [147]. See supra note 42 and accompanying text.

 [148]. Cf. Birger Dölling, Der fliegende Gerichtsstand im Presserecht: Spielball der Interessen?, 68 NJW 113, 126 (2015).

 [149]. Jürgens, Abgestürzte Gerichtsstände, supra note 146, at 3065; Constantin van Lijnden, Fliegende Richter, Zeit Online (July 14, 2016, 3:01 AM), http://www.zeit.de/2016/28/presserecht-justiz-verfahren-gerichtsstand.

 [150]. See generally Jürgens, Abgestürzte Gerichtsstände, supra note 146 (arguing that forum shopping has resulted in a situation where most press law cases are heard by judges favoring plaintiffs); Uwe Jürgens, Turbulenzen im Presseprozessrecht: Der rechtswidrige Standardverzicht auf mündliche Verhandlungen im einstweiligen Rechtsschutz, 19 Kommunikation & Recht [K&R] 17 (2016) [hereinafter Jürgens, Turbulenzen im Presseprozessrecht] (arguing that flexible venue rules have led to the prevalence of ex parte injunctions in press law). For discussions of alternative viewpoints, see generally Dölling, supra note 148 (arguing that there is neither evidence for widespread forum shopping in press law, nor for the assertion that the most important press law courts systematically favored plaintiffs); Ralf Höcker & Lucas Brost, Kompetenz zahlt sich aus: Zur Notwendigkeit des fliegenden Gerichtsstandes im Nischenrechtsgebiet Presserecht, 5 IP-Rechtsberater [IPRB] 138 (2015) (questioning the evidence supporting forum shopping in press law, and arguing that flexible venue rules are required to maintain the quality of press law chambers).

 [151]. Dieter Leuze, in Berliner Kommentar zum Grundgesetz C Art. 101, ¶ 17 (Karl-Heinrich Friauf & Wolfram Höfling eds., 2018).

 [152]. Deutsche Vereinigung für gewerblichen Rechtsschutz und Urheberrecht, Überlegungen des Bundesministeriums der Justiz zu Änderungen im Recht der einstweiligen Verfügung, 111 GRUR 564 (2009); see also Dölling, supra note 148, at 125.

 [153]. Regierungsentwurf [Cabinet Draft], BT 17/13057, 31971, https://dip21.bundestag.de/dip21
/btd/17/130/1713057.pdf. The Federal Parliament (Bundestag), had restricted venue in copyright cases in which an individual was sued for non-commercial use of protected works. See Urheberrechtsgesetz [UrhG] [Copyright Act], Sept. 9. 1965, BGBl. I at 1273, as amended by Act of Nov. 28, 2018 BGBl I at 2014, § 104a, translation at http://www.gesetze-im-internet.de/englisch_urhg/index.html.

 [154]. Stenografischer Bericht, BT 17/250, 146, http://dip21.bundestag.de/dip21/btp/17/17250_ ORIG.pdf; Beschlussempfehlung des Rechtsausschusses, BT 17/14192, 5, https://dip21.bundestag.de/ dip21/btd/17/141/1714192.pdf.

 [155]. Beschluss der Ministerinnen und Minister: “Fliegender Gerichtsstand 14 Absatz 2 Satz 1 UWG), Konferenz der Justizministerinnen und Justizminister (Nov. 17, 2016), https://www.justiz.nrw.de/JM/jumiko/beschluesse/2016/Herbstkonferenz-2016/top7_-_fliegender
_gerichtsstand_herbstkonferenz.pdf.

 [156].                In 2018, the Federal Ministry of Justice issued a legislative proposal which would restrict venue choice in unfair competition law, but not in press law. Referentenentwurf [Ministerial Draft], Entwurf eines Gesetzes zur Stärkung des fairen Wettbewerbs (2018), https://www.bmjv.de
/SharedDocs/Gesetzgebungsverfahren/Dokumente/RefE_fairerWettbewerb.pdf?__blob=publicationFile&v=1.

 [157].                See infra notes 16870 and accompanying text.

 [158]. Preliminary injunctions do not automatically expire. Instead, the defendant can obtain a court order asking the plaintiff to open principal proceedings. If the plaintiff fails to do so, the court has to revoke the preliminary injunction. German Code of Civil Procedure, supra note 12, §§ 926, 936. Different from injunctions and orders requiring the press outlet to publish a reply, judgments for damages and judgments ordering the press outlet to issue a correction can normally not be issued in preliminary proceedings, but require the court to hold principal proceedings. See Ingo Drescher, in Münchener Kommentar zur Zivilprozessordnung, supra note 68, § 935, ¶¶ 65–67; Lutz Haertlein, in Gesamtes Recht der Zwangsvollstreckung § 935, ¶ 27 (Johann Kindl et al. eds., 3d ed. 2016).

 [159]. German Code of Civil Procedure, supra note 12, § 937(2).

 [160]. Wolf-Dietrich Walker, Der einstweilige Rechtsschutz im Zivilprozess und im arbeitsgerichtlichen Verfahren 280 (1993); Wolf-Dietrich Walker, Die Schutzschrift und das elektronische Schutzschriftenregister nach §§ 945a, 945b ZPO, in Rechtslage – Rechtserkenntnis Rechtsdurchsetzung: Festschrift für Eberhard Schilken zum 70. Geburtstag 815, at 815 (Caroline Meller-Hannich et al. eds., 2015) [hereinafter Walker, Die Schutzschrift].

 [161]. See Bundesverfassungsgericht [BVerfG] [Federal Constitutional Court], June 6, 2017, 1 BvQ 16/17, 1 BvQ 17/17, 1 BvR 764/17, 1 BvR 770/17, ¶ 2, http://www.bundesverfassungsgericht.de/ SharedDocs/Entscheidungen/DE/2017/06/qk20170606_1bvq001617.html; Rafael Buschmann et al., Bitte bellen Sie leise, Der Spiegel, Aug. 17, 2017, at 99, 99100.

 [162]. Interview with attorney (May 17, 2017).

 [163]. Dölling, supra note 148, at 127; Jürgens, Abgestürzte Gerichtsstände, supra note 146, at 3064; Jürgen Dahlkamp & Barbara Schmid, Köln nimmt alles, Der Spiegel, Oct. 13, 2014, at 150, 150; Höcker & Brost, supra note 150, at 139. Note that some of our interviewees disputed this account.

 [164]. For corroboration of these accounts, see BVerfG, Sept. 30, 2018, 1 BvQ 1783/17, ¶ 8, https://www.bundesverfassungsgericht.de/SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr178317.html; BVerfG, Sept. 30, 2018, 1 BvQ 2421/17, ¶ 21, https://www.bundesverfassungs
gericht.de/SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr242117.html.

 [165]. See Otto Teplitzky, Gewohnheitsunrecht? Anmerkungen zum Einfluss der normativen Kraft des Faktischen auf die einstweilige Unterlassungsverfügung, in Festschrift für Joachim Bornkamm zum 65. Geburtstag 1073, 1086 (Wolfgang Büscher et al. eds., 2014) (discussing ex parte injunctions in general).

 [166]. Both the German constitution and the grant defendants in Germany the opportunity to be heard. Grundgesetz [GG] [Basic Law], art. 103, § 1; European Convention on Human Rights, art. 6, § 1; see also Teplitzky, supra note 165, at 1087.

 [167]. See Walker, Die Schutzschrift, supra note 160, at 816.

 [168].                BVerfG, Sept. 30, 2018, 1 BvQ 1783/17, https://www.bundesverfassungsgericht.de/
SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr178317.html; BVerfG, Sept. 30, 2018, 1 BvQ 2421/17, https://www.bundesverfassungsgericht.de/SharedDocs/Entscheidungen/DE/2018/09/
rk20180930_1bvr242117.html.

 [169].                BverfG, Sept. 30, 2018, 1 BvQ 1783/17, ¶ 18–19, https://www.bundesverfassungs
gericht.de/SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr178317.html; BverfG, Sept. 30, 2018, 1 BvQ 2421/17, ¶ 31–32, https://www.bundesverfassungsgericht.de/SharedDocs/Entscheidungen/
DE/2018/09/rk20180930_1bvr242117.html.

 [170].                BverfG, 1 BvQ 2421/17, Sept. 30, 2018, 36, https://www.bundesverfassungsgericht.de/
SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr242117.html.

 [171]. See German Code of Civil Procedure, supra note 12, § 945a. Traditionally, such a protection letter had to be filed with every court individually. Since 2016, there is a central registry with which protection letters can be filed.

 [172]. Ingo Drescher, in Münchener Kommentar zur Zivilprozessordnung, supra note 68, § 945a, ¶ 6.

 [173]. In line with this argument, the Federal Constitutional Court in its aforementioned rulings decided that the possibility of filing a protection letter renders it unnecessary to involve the defendant in preliminary ruling proceedings only under certain conditions. Most importantly, the Court requires a cease and desist letter whose content should be basically identical with the motion for preliminary injunction. BVerfG, Sept. 30, 2018, 1 BvQ 1783/17, ¶¶ 22–23, https://www.bundesverfassungs
gericht.de/SharedDocs/Entscheidungen/DE/2018/09/rk20180930_1bvr178317.html; BVerfG, Sept. 30, 2018, 1 BvQ 2421/17, ¶¶ 34–35, https://www.bundesverfassungsgericht.de/SharedDocs/Entscheidungen/
DE/2018/09/rk20180930_1bvr242117.html; see also Teplitzky, supra note 165, at 1088.

 [174]. Jörg Soehring, Presserecht: Recherche, Darstellung, Haftung im Recht der Presse, des Rundfunks und der neuen Medien 745 (Jörg Soehring & Verena Hoene eds., 5th ed. 2013); cf. Rolf Nikolas Danckwerts, Die Entscheidung über den Eilantrag, 110 GRUR 763, 764–65 (2008) (arguing that hearing the defendant cannot change the outcome of a proceeding if the facts of the case are undisputed).

 [175]. German Code of Civil Procedure, supra note 12, §§ 92425, 936.

 [176]. Id. § 542(2).

 [177]. Id. §§ 926, 936. If the plaintiff fails to do so within the time period set by the court, the preliminary injunction has to be rescinded by the court. Decisions in principal proceedings follow general rules.

 [178]. See BVerfG, June 6, 2017, 1 BvQ 16/17, 1 BvQ 17/17, 1 BvR 764/17, 1 BvR 770/17, ¶ 7, http://www.bundesverfassungsgericht.de/SharedDocs/Entscheidungen/DE/2017/06/qk20170606_1bvq001617.html.

 [179]. German Code of Civil Procedure, supra note 12, §§ 921, 945. The purpose of the bond is solely to serve as a security for a defendant’s claim for damages in case the preliminary injunction is later revoked.

 [180]. Jürgens, Turbulenzen im Presseprozessrecht, supra note 150, at 10.

 [181]. Haertlein, supra note 158, § 935, ¶ 27. German case law is split on whether courts can and will grant a motion for a preliminary injunction which had been previously filed in another court. See, e.g., Kammergericht Berlin [KG Berlin] [Court of Appeals Berlin] Oct. 10, 2016, GRUR 128, at ¶ 3. It is unclear whether plaintiffs are obliged to disclose to the second court that they had requested a preliminary injunction from another court, and courts will usually not inquire about this. Judges might only raise this issue if the circumstances of the case make it apparent that the request had been filed elsewhere before, for example if the request was filed several weeks after the cease and desist letter was sent. See Helmut Köhler, in Gesetz gegen den unlauteren Wettbewerb: Kommentar § 12 UWG ¶ 3.16a (Helmut Köhler et al. eds., 36th ed. 2018); Otto Teplitzky, Unzulässiges forum-“hopping nach gerichtlichen Hinweisen, 62 WRP 917, 917–19 (2016).

 [182]. One exception to this rule is the regional court in Hamburg, which is reported to sometimes take more time than other major courts.

 [183]. Interview with attorney (May 17, 2017).

 [184]. Interview with attorney (Mar. 30, 2017).

 [185]. Apparently, judges in Frankfurt currently apply rules concerning the standard of proof regarding the truth of reports on criminal acts by individuals in a way that is particularly advantageous for individuals.

 [186]. Gerhard Wagner, Rechtsstandort Deutschland im Wettbewerb 37–38 (2017).

 [187]. Jürgens, Abgestürzte Gerichtsstände, supra note 146, at 3065.

 [188]. See also Sascha Sajuntz, Die Entwicklung des Presse- und Äusserungsrechts in den Jahren 2012/2013, 67 NJW 25, 30 (2014).

 [189]. One interviewee even suggested that courts deliberately set the amount of controversy in a way that makes it impossible for parties to appeal decisions by the regional courts of appeals to the Federal Court of Justice. This claim seems dubious given the power of the Federal Court of Justice to review the determination of the amount in controversy of the lower courts. See Ingo Saenger, in Zivilprozessordnung EGZPO § 26, ¶ 10 (Ingo Saenger ed., 7th ed. 2017); see also BGH [Federal Court of Justice], May 15, 2014, GRUR-RS 11248, 2014.

 [190]. Rechtsanwaltsvergütungsgesetz [RVG] [Act on the Remuneration of Attorneys], May 5, 2004, BGBl. I at 718, 788, as amended by Act of Dec. 17, 2018 BGBl I at 2573, § 2, translation at http://www.gesetze-im-internet.de/englisch_rvg/index.html.

 [191]. On speed, see supra Section IV.B.3.

 [192]. See Jürgens, Abgestürzte Gerichtsstände, supra note 146, at 3066.

 [193]. Interview with attorney (Dec. 20, 2017).

 [194]. Interview with attorney (Feb. 26, 2018).

 [195]. Interview with attorney (Sept. 21, 2016).

 [196]. Id.

 [197]. Uwe Jürgens estimates that the revenues from press law cases heard by one of the major regional courts amount to around one million euros per year. Jürgens, Abgestürzte Gerichtsstände, supra note 146, at 3066.

 [198]. Case C-453/99, Courage v. Crehan, 2001 E.C.R. I-6314. On the impact of this decision, see Ioannis Lianos et al., Damages Claims for the Infringement of EU Competition Law ¶¶ 2.13–.16 (2015).

 [199]. Council Regulation (EU) 1/2003, of Dec. 16, 2002 on the Implementation of the Rules on Competition Laid Down in Articles 81 and 82 of the Treaty, 2003 O.J. (L 1) 1.

 [200]. Commision of the European Communities, White Paper on Damages Actions for Breach of the EC Antitrust Rules, COM (2008) 165 final (Apr. 2, 2008). On the development of private enforcement of European antitrust law in general, see David Ashton, Competition Damages Actions in the EU 37, ¶¶ 0.08–.25 (2d ed. 2018); Assimakis Komninos, Private Enforcement in the EU with Emphasis on Damages Actions, in Handbook on European Competition Law: Enforcement and Procedure 228, 23447 (2013); Lianos et al., supra note 198, ¶¶ 2.10–.45.

 [201]. Consolidated Version of the Treaty on European Union and the Treaty on the Functioning of the European Union art. 101, May 9, 2008, 2008 O.J. (C 115) 88; Gesetz gegen Wettbewerbsbeschränkungen [GWB] [Act against Restraints of Competition], June 26, 2013, BGBl. I at 1750, 3245, as amended by Act of July 12, 2018 BGBl I at 1151, § 1, translation at https://www.gesetze-im-internet.de/englisch_gwb/index.html [hereinafter German Act Against Restraints of Competition].

 [202]. The damage caused by a cartel is typically widely spread across a number of direct and indirect customers. U.S. civil procedure allows customers to file a class action lawsuit in order to recoup such damages in an effective manner. German civil procedure allows neither class actions nor contingency fees. As a result, German plaintiffs have relied on other mechanisms. In one model, a third party buys claims from different consumers, sues the cartel members in its own name and then distributes parts of the damages to the consumers. A company offering this model is Cartel Damage Claims, based in Brussels. In another model, a large consumer creates its own subsidiary to investigate systematically all relationships to its suppliers for potential antitrust violations. The subsidiary then sues these suppliers and collects damages. The German railway company Deutsche Bahn did this with a special unit called CRK4 and a subsidiary called Barnsdale Cartel Damage Solutions to enforce damage claims against numerous cartels the company had been harmed by. See Ashton, supra note 200, at 28283, ¶¶ 11.01–.07, 33234, ¶¶ 11.201–.204; Fabian Stancke, Rechtliche Rahmenbedingungen kartellrechtlicher Massenklagen, 68 Wirtschaft und Wettbewerb [WUW] 59, 59 (2018).

 [203]. These courts are based in Berlin, Bremen, Cologne, Dortmund, Düsseldorf, Erfurt, Frankfurt a.M., Gera, Hamburg, Hannover, Kassel, Kiel, Leipzig, Magdeburg, Mainz, Mannheim, Meiningen, Mühlhausen, Munich, Nuremburg-Fürth, Potsdam, Rostock, Saarbrücken, and Stuttgart.

 [204]. German Code of Civil Procedure, supra note 12, §§ 17, 32.

 [205]. See supra note 11.

 [206]. German Act against Restraints of Competition, supra note 201, § 33b.

 [207]. Ashton, supra note 200, at 19091, ¶¶ 7.34–.36; Alexander Fritzsche et al., Die Praxis der privaten Kartellrechtsdurchsetzung in Deutschland, 4 Neue Zeitschrift für Kartellrecht [NZKart] 412, 418–20 (2016); Christian Kersting, Kartellschadensersatz: Haftungstatbestand, Bindungswirkung, Schadensabwälzung, in Die 9. GWB-Novelle 133, 142 (Christian Kersting & Rupprecht Podszun eds., 2017).

 [208]. See, e.g., Oberlandesgericht Jena [OLG Jena] [Court of Appeals Jena], Feb. 22, 2017, 67 WUW 203.

 [209]. Andreas Weitbrecht, Kartellschadensersatz 2017, 6 NZKart 106, 111 n.67 (2018), lists Dortmund, Hannover, and Mannheim as particularly active courts.

 [210]. Landgericht Dortmund [LG Dortmund] [Regional Court of Dortmund], Apr. 1, 2004, 54 WUW 1182; see Thomas Thiede & Tim Träbing, Praxis des Anscheinsbeweises im Kartellschadensersatzrecht ein Rechtsprechungsbericht, 4 NZKart 422, 424 (2016).

 [211]. For an overview, see Thiede & Träbing, supra note 210, at 424–27.

 [212]. Landgericht Mannheim [LG Mannheim] [Regional Court of Mannheim], May 4, 2012, 12 NJOZ 1635.

 [213]. Interview with attorney (Mar. 18, 2016).

 [214]. Id.

 [215]. Geschäftsverteilung des Landgerichts Mannheim für das Geschäftsjahr 2018, Regional Court of Mannheim (2017) (on file with the authors).

 [216]. Press release,                                                         Regional Court of Mannheim, Landgericht Mannheim baut wirtschaftsrechtlichen Schwerpunkt weiter aus: Zusätzliche Kammer für Kartellverfahren eingerichtet, (Sept. 24, 2018), http://www.landgericht-mannheim.de/pb/,Lde/Startseite/Aktuelles/Pressemitteilung+
vom+24_09_2018/?LISTPAGE=1167835.

 [217]. A recent study of published court decisions confirms such differences between courts: of the fifty-one first instance court decisions on antitrust follow-on lawsuits that were decided between 2003 and 2017, eleven cases were decided by the regional court in Mannheim, ten in Dortmund and six in Berlin. Lukas Rengier, Kartellschadenersatz in Deutschland die ersten 15 Jahre in Zahlen und Lehren für die Zukunft, 68 WUW 613, 619 (2018).

 [218]. BGH [Federal Court of Justice], Dec. 11, 2018, 7 NZKart 101 (2019) (holding that cartels do not have sufficient typical patterns that are necessary to create a prima facie evidentiary rule under German civil procedure law); see also Tilmann Hertel et al., Anscheinsbeweis Adieu – Gezeitenwechsel für den Schadensnachweis bei Follow-on Klagen, 7 NZKart 86, 86–88 (2019); Stefan Rützel, Abschied vom “doppelten Anscheinsbeweis”, 69 WUW 130, 131–32 (2019) (both discussing the implications of the decision of the Federal Court of Justice for future antitrust litigation).

 [219].               In 2017, the German legislature enacted a statutory burden shifting in a reform of the Act against Restraints of Competition. According to Section 33a(2) of the revised Act, there is now a rebuttable presumption that a cartel caused damage to the plaintiff, thereby overcoming challenges (2) and (3) mentioned before. While the plaintiff can benefit from this burden shifting as far as the existence of a damage is concerned, she still has to prove the damage amount (challenge (4)). The decision of the Federal Court of Justice dealt with a case before the statutory burden shifting came into force. It is too early to tell how both the statutory burden shifting and the decision by the Federal Court of Justice will influence decision patterns by lower courts in antitrust follow-on lawsuits. For example, in a decision from January 2019,, the Court of Appeals of Düsseldorf did not follow the (higher) Federal Court of Justice, but stuck to its own case law and applied a prima facie evidentiary presumption in an antitrust follow-on lawsuit, OLG Düsseldorf [Court of Appeals Düsseldorf], Jan. 1, 2019, 7 NZKart 157 (2019).

 [220]. Interview with attorney (Mar. 18, 2016).

 [221]. Rengier, supra note 217, at 618.

 [222]. Interviewees have told us of a case featuring up to ten defendants and fifteen third-parties (Streitverkündete), in which all of these parties replied to the plaintiff’s arguments separately; and another case involving six cartel members and thirty-two injured firms with ninety-six production facilities in fourteen E.U. member states. In yet another case, the plaintiffs submitted 320,000 receipts which were printed as hard copies twelves times and were delivered to the court in 475 folders. Generally, it is not uncommon for judges in these cases to read hundreds of pages of briefs.

 [223]. Gerhard Klumpe & Thomas Thiede, Ergänzende Überlegungen zu Lukas Rengier, Kartellschadensersatz in Deutschland, 7 NZKart 136, 138 (2019).

 [224]. Interview with attorney (Feb. 18, 2017).

 [225]. Note that, while the amount in controversy in follow-on lawsuits can be in the hundreds of millions of euros or even higher, the German Act on Court Fees (Gerichtskostengesetz) puts a statutory cap on the amount in controversy at €30 million. German Act on Court Fees, supra note 124, § 39(2). This severely limits the ability of courts to generate court fees in very large proceedings.

 [226]. Statistische Ämter des Bundes und der Länder, Ergebnisse der Steuerstatistiken 33 (2014).

 [227]. Theoretically, one could imagine that courts in states without such industry could step in and position themselves as vigilant antitrust enforcers. However, at least for pan-European cartels, it may be impossible for plaintiffs to file a lawsuit in such states if no cartel member has a production facility there. As one of our interview partners noted, the prevalence of the Stuttgart and Dortmund district courts in antitrust follow-on litigation may also result from the fact that two large cartel members (Daimler and ThyssenKrupp) are headquartered in the respective districts. For venue rules in pan-European damage claims cases, see infra note 232.

 [228]. See Melissa Eddy & Jack Ewing, As Europe Sours on Diesel, Germany Groups Fights to Save It, N.Y. Times Int’l Edition, Aug. 2, 2017, at B2 (“Vehicles are Germany’s single most important export product and, in many parts of the world, the most visible symbol of German engineering prowess. Within the country, BMWs, Mercedes-Benzes and Porsches are a source of considerable pride and an essential part of the postwar national self-image.”); Jack Ewing, As German Election Looms, Politicians Face Voters’ Wrath for Ties to Carmakers, N.Y. Times Int’l Edition, Sept. 13, 2017, at B1 (“Sometimes it is hard to tell where the German government ends and the auto industry begins. . . . For decades, the German government has been a crucial ally for carmakers, operating as a de facto lobbyist for the industry”).

 [229]. Exceptions to this rule include the Court of Justice of the European Union, which has the power to determine the interpretation of E.U. law.

 [230]. The current regulation is the recast Brussels I Regulation. Regulation No. 1215/2012, of the European Parliament and of the Council of Dec. 12, 2012 on Jurisdiction and the Recognition and Enforcement of Judgments in Civil and Commercial Matters, 2012 O.J. (L 351) 1 (EU). In its CDC decision, the European Court of Justice dealt with the original Brussel I Regulation from 2001. Case C-352/13, Cartel Damage Claims (CDC) Hydrogen Peroxide SA v. Evonik Degussa GmbH, 2015 ECR I-335 ¶¶ 36. The European Court of Justice did not allow follow-on lawsuits in E.U. member states where the cartel had a negative impact on prices, but no cartel member had a domicile.

 [231]. Regulation No. 864/2007, of the European Parliament and of the Council of July 11, 2007 on the Law Applicable to Non-contractual Obligations (Rome II), art. 6, 2007 O.J. (L 199) 40, 44 (EC).

 [232]. This only holds true if the plaintiff sues cartel members at a location where one member is domiciled. If the plaintiff sues cartel members at the location where they formed the cartel, the plaintiff can only claim the damages that occurred in the respective E.U. member state. This explains the importance of identifying an anchor defendant in pan-European follow-on cases. See Case C-352/13, 2015 ECR I-335 ¶¶ 33, 54 ; Jens Adolphsen & Frederik Möller, Bestimmung des anwendbaren Rechts, in Kartellverfahren und Kartellprozess: Handbuch 1173, 1184, ¶ 38 (Hans-Georg Kamann et al. eds., 2017); Wolfgang Wurmnest, Forum Shopping bei Kartellschadensersatzklagen und die Kartellschadensersatzrichtlinie, 5 NZKart 2, 5 (2017) [hereinafter Wurmnest, Forum Shopping]; Wolfgang Wurmnest, International Jurisdiction in Competition Damages Cases Under the Brussels I Regulation: CDC Hydrogen Peroxide, 53 Common Mkt. L. Rev. 225, 227–28 (2016). If none of the cartel members are domiciled in the country where the plaintiff intends to sue, the plaintiff may still sue a subsidiary of a cartel member and argue that the subsidiary participated in the implementation of the cartel, thereby using it as an anchor defendant. See Provimi Ltd. v. Agentis Animal Nutrition SA [2003] EWHC (QB) 961, 7 Eur. L. Rep. 517, 517–18 (Eng.).

 [233]. This does not mean that a German court dealing with a follow-on lawsuit will see courts from other E.U. member states as a direct competitor. As one of our interviewees noted, the typical German judge may still not see or even care that there is a competition emerging from other European countries.

 [234]. Understandably, a German attorney may not have an interest to advise his client to sue in another country, as a local attorney would take over this case. One of the attorney interviewees explained that “[i]f the lawsuit is not filed in Germany, then it will probably be filed without us. So it is again a case without a wild boar. And parties without a wild boar are less good than parties with a wild boar.” Interview with attorney (Jan. 31, 2017).

 [235]. Matthijs Kuipers et al., Action for Damages in the Netherlands, the United Kingdom, and Germany, 6 J. Eur. Competition L. & Prac. 129, 129–30 (2015). Similar to Germany, very few Dutch or U.K. decisions exist in which a court did not only rule on the general question of whether cartel members are liable for follow-on damages, but also awarded specific damages. Our interviewees also reported of strategies by potential defendants to block lawsuits by defendants in particular jurisdictions (for example, in the costly United Kingdom or in Israel) through filing an action for declaratory judgment that the cartel member is not liable in a country such as Germany or Italy.

 [236]. As nearly all cartel contracts and related documents are written in English, the Dutch, and the British court systems are at an important advantage. Many German courts will not accept expert opinions and other supplementary material in English. In the Netherlands, documents can be submitted in English (as well as Dutch, French or German). As to information technology, as reported above, many German courts still require documents to be filed on paper. In the Netherlands, plaintiffs can submit documents on DVD, and briefs are usually no longer than seventy pages.

 [237]. A plaintiff in a German court has to pay court fees in advance. In one well-known German antitrust follow-on lawsuit, the plaintiffs had to deposit €2.3 million in court fees in advance. See Wurmnest, Forum Shopping, supra note 232, at 7 n.86.

 [238]. As described supra text accompanying note 232, it is beneficial for plaintiffs to sue the cartel in a country in which one of the cartel members is domiciled, as the plaintiff can the claim European Unionwide damages in that proceeding. This may seem to count against the Netherlands, as fewer companies are domiciled in the Netherlands than in, for example, Germany. However, as mentioned supra note 232, it may be to sufficient to use the Dutch subsidiary of a cartel member as an anchor defendant, thereby overcoming this apparent hurdle.

 [239]. Arianna Andreangeli, The Consequences of Brexit for Competition Litigation: An End to a “Success Story”?, 38 Eur. Competition L. Rev. 228, 22829 (2017) (arguing that Brexit may reduce the appeal of U.K. courts).

 [240]. Ashton, supra note 200, at 8890, ¶¶ 4.12–.17.

 [241]. Ashton, supra note 200, at 31020, ¶¶ 11.116–.152.

 [242]. Andreangeli, supra note 239, at 233.

 [243]. See infra Section VII.B.4.

 [244]. See Graham & Van Zeebroeck, supra note 30, at 678.

 [245]. See Cotter, supra note 46, at 233–36. See generally Graham & Van Zeebroeck, supra note 30 (discussing patent litigation in Europe). In fact, patent litigation also has a global dimension. See Eun-Joo Min & Johannes Wichard, Cross-Border Intellectual Property Enforcement, in The Oxford Handbook of Intellectual Property Law 687, 687–89 (Rochelle C. Dreyfuss & Justine Pila eds., 2018). For recent papers that explore firm strategies in the “global patent wars” of the smartphone industry and related fields, see Stuart Graham & Saurabh Vishnubhakat, Of Smart Phone Wars and Software Patents, 27 J. Econ. Persp. 67, 67–71 (2013); Paik & Zhu, supra note 30, at 1398; Michael Elmer & Stacy Lewis, Where to Win: Patent-Friendly Courts Revealed, Managing Intell. Prop., Oct. 2010, at 1, 12.

 [246]. See supra Section VI.A.

 [247]. While many litigated patents originate from a patent issued by the European Patent Office, these patents are transformed into national patents after issuance. Therefore, technically, a patentee requesting an injunction in two different jurisdictions has to base his claim on the infringement of two separate national patents. In the late 1990s, mostly in the Netherlands, national courts nevertheless started issuing so-called cross-border injunctions covering multiple jurisdictions. This practice was perceived by some as an attempt to attract more cases to Dutch courts. It was effectively stopped by a judgment of the CJEU determining that national courts have exclusive jurisdiction to decide cases in which the validity of a national patent is at issue. C-4/03, Gesellschaft für Antriebstechnik mbH & Co. KG v. Lamellen und Kupplungsbau Beteiligungs KG, 2006 E.C.R. I-06523 ¶ 47. Later, the CJEU narrowed down this effective ban by allowing cross-border injunctions in preliminary proceedings. C-616/10, Solvay SA v. Honeywell Fluorine Products Europe BV and Others, 2012 E.C.R. I-06509 ¶ 56; see also Cotter, supra note 46, at 250–55; Tilman Müller-Stoy & Jörg Whal, The European Union: Jurisdiction, Cross-Border-Cases, Enforcement Directive and Unified Patent Court, in Patent Enforcement Worldwide: Writings in Honour of Dieter Stauder 57, 65 (Christopher Heath ed., 3d ed. 2015); Graham & Van Zeebroeck, supra note 30, at 674. In addition, U.K. courts offer potential defendants the ability to obtain a European Union–wide declaratory judgment that a product does not infringe a certain patent. Because U.K. courts lack jurisdiction to rule on the validity of patents in other jurisdictions, these defendants have to forego the opportunity to attack the validity of the patent. For additional discussion on this topic, see Gregory Bacon & Katie Rooth, Justiciability and Litigation of Foreign Patents in the English Courts, 12 J. Intell. Prop. L. & Prac. 851, 858 (2017). Once the so-called Unitary Patent and the Unified Patent Court are introduced on a European level, local divisions of the Unified Patent Court may be in much more direct competition than the current national patent courts in Europe. See Müller-Stoy & Whal, supra, at 69.

 [248]. There are some cases in which courts in different jurisdictions might nevertheless be regarded as viable alternatives. Plaintiffs might use injunctions issued in one jurisdiction as a bargaining chip in negotiations aimed at settling patent disputes across jurisdictions. Then, parties might aim for being the first to obtain an injunction in their favor irrespective of whether this injunction covers all the jurisdictions where the battle is playing out. Also, a country like Germany may be so important economically that it just may not make sense to sell a product in the European Union if a court has enjoined the sale of the product in Germany, even if the producer would still be allowed to sell the product in other European countries.

 [249]. See Cotter, supra note 46, at 235–36 (citing studies estimating that, as of 2000, the average cost of proceeding to judgment in a court of first instance were highest in the U.K. (between €150,000 for a small-medium case and €1.5 million for a larger case) and lowest in France (between €50,000 and €200,000), with Germany (between €50,000 and €250,000) and the Netherlands (between €60,000 and €200,000) lying in the middle).

 [250]. On the advantages and disadvantages of various European jurisdictions for patent infringement cases, see generally Luke McDonagh, European Patent Litigation in the Shadow of the Unified Patent Court (2016).

 [251]. Cotter, supra note 46, at 229; Herr & Grunwald, supra note 36, at 44; Malte Köllner et al., supra note 84, at 9–11 (analyzing patent litigation data from Darts IP suggesting that German courts stay proceedings at a lower rate than Dutch, French, or U.K. courts).

 [252]. See TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514, 1520 (2017).

 [253]. An important example is the decision whether to stay a proceeding, see supra Section III.B.4.

 [254].  Exceptions to this rule are limited to gross violations of procedural norms. See German Code of Civil Procedure, supra note 12, § 547.

 [255]. Klerman & Reilly, supra note 1, at 245.

 [256]. On the role of specialization in competition among courts, see J. Jonas Anderson, Court Competition for Patent Cases, 163 U. Pa. L. Rev. 631, 636–37 (2015); Klerman & Reilly, supra note 1, at 303.

 [257].                Note that, even if there are multiple chambers with a certain specialization in a particular court, a potential plaintiff might be able to predict or even influence which chamber is going to hear a specific case. One of our interview partners indicated that applicable strategies include manipulating the order of plaintiffs in a multi-plaintiff lawsuit (if cases are allocated to chambers according to plaintiff names); withdrawing a lawsuit if the case ends up with the “wrong” chambers; or getting background information from the secretary of the court chamber. We were, however, unable to ascertain the existence and prevalence of such troubling attorney strategies, and doubt whether they play a major role.