Asymmetric cross-citations in private law: An empirical study of 28 supreme courts in the EU

A. ‘True followers’

The first category of foreign supreme court citations refers to the category of ‘true followers’. As noted above, these countries fulfil three criteria. First, in all of their judgments delivered in the examined time period, they cited foreign supreme courts at least 10 times (absolute number). Second, these citations made up at least 1% of the total number of judgments delivered (relative number). Third, more than 66% of the citations referred to a single foreign supreme court, ³⁸ that is to say, these countries closely follow the jurisprudence of one particular foreign supreme court. In the present case, these countries are Austria, Cyprus, Ireland and Luxembourg. As shown in Figure 2, the Austrian Supreme Court cited most frequently the German Supreme Court (1044 out of 1068 citations). For the supreme courts of Cyprus and Ireland, it was the Supreme Court of the United Kingdom (223 out of 229 citations for Cyprus and 192 out of 194 citations for Ireland), whereas the Luxembourg court referred to the French court the most regularly (45 out of 57 citations).

OPEN IN VIEWER

Figure 2.

Cross-citations of the supreme courts of Austria, Cyprus, Ireland and Luxembourg. Explanatory note: The scale of the chart refers to the relative frequency of cross-citations while the numbers indicated in brackets in each column refers to absolute numbers of cross-citations of each court.

Already at a first glance there seems to exist a plausible explanation to this finding. All four ‘true followers’ are relatively small jurisdictions with historic ties to neighbouring larger jurisdictions whose legal developments they may traditionally have reasons to follow. This does not mean that amongst these countries the level of the foreign influence is identical. In particular, the information about the relative frequency in Figure 2 shows that the attachment of Ireland to the UK far exceeds those of the other three countries (in line with prior research identifying a high degree of such influence in the case law of the Irish Supreme Court). ³⁹

For each of the four countries, the following more specific explanations can be added: German legal thinking is known to have had a strong influence on many legal systems in Central and Eastern Europe, including Austria. ⁴⁰ While the Austrian Civil Code (ABGB) predates the German Civil Code (BGB), nowadays the interpretation of the ABGB is often inspired by the interpretation of the BGB, similarly to other areas of private law. ⁴¹ Our observation that the Austrian supreme court (Oberster Gerichtshof) mostly cites the German court comes as no surprise in light of this. By comparison, citations to other foreign courts are significantly less frequent: Italy (9 citations), France (7 citations), the Netherlands (3 citations), the UK (3 citations), Ireland (1 citations) and Luxembourg (1 citations).

The Cypriot legal system, often referred to as a mixed system, retains essential elements of the English common law system, due to the UK’s relatively recent foreign rule in Cyprus, which ended only in 1960. To give an example, while Cyprus does not use the jury system for civil or criminal trials, it applies the principle of stare decisis as a result of its common law origins. ⁴² However, as the Cypriot legal system is mixed, it is also influenced by further legal traditions. According to Symeonides, the two main areas of law where Cyprus has typically drawn from non-English sources (that is, Greek, French and more traditionally Roman-Byzantine and Ottoman law) are administrative law and constitutional law. ⁴³ This explains our findings regarding significant private law references by the Cypriot supreme court (ανώτατο Δικαστήριο) to the UK highest court, as Cypriot private law has heavily been influenced by common law. ⁴⁴ By comparison, the Greek influence is modest in this regard with just 5 of the remaining cross-citations.

Ireland acquired independence from the United Kingdom in 1922, inheriting its common law system though showing some variations to it. ⁴⁵ For example, Ireland’s written constitution is something the UK still lacks. Upon achieving independence, Ireland adopted the bulk of the pre-existing body of British law, subjecting it to review against the standards of the Irish Constitution. ⁴⁶ In this sense, pre-1922 British case law still forms part of the Irish legal system. The Irish Supreme Court often refers to these older pre-independence precedents as well as more modern UK cases, in particular in matters of private law. ⁴⁷ Given this close historic relationship of the two legal systems, it comes as no surprise to see the Irish Supreme Court’s careful observance of UK case law.

The final country of this category is Luxembourg. The Luxembourgish legal system has been influenced to a great extent by its neighbouring countries, such as France (mostly its civil code and civil procedure) ⁴⁸ and Belgium (for its commercial and criminal codes, as well as its constitutional system). ⁴⁹ Thus, it seems logical that the overwhelming majority of references from the Luxembourg Supreme Court (Cour de Cassation) ⁵⁰ concern the French Supreme Court (45 citations), and – to a lesser degree – the Belgian Supreme Court (12 citations).

As noted above, ‘true followers’ tend to be small jurisdictions whose positive law follows the law of another country. Apart from the historic ties between these countries, there is another factor which is worth considering: language. At least to some extent, all these country pairs share the same official language. For Austria and Germany, it is German; for Ireland and the UK, it is English (even though the first official language of Ireland is Irish, English has an equal status to Irish ⁵¹ ). France and Luxembourg share the French language, which remains the main legal language in Luxembourg, despite Luxembourgish also being an official language. ⁵² The only partial outlier is Cyprus, given that in 1960 Greek and Turkish replaced English as the official language of the country. ⁵³ Nevertheless, it took three decades for Greek to replace English as the language of the judiciary, and even today English terms and sources are used in court proceedings. ⁵⁴ This explains why the overwhelming majority of Cypriote court references cite UK and not Greek case law.

Finally, it is noteworthy that jurisdictions falling into the category of ‘true followers’ belong to diverse legal families. Austria and Luxembourg are civil law countries, belonging to the Germanistic and Romanistic legal branch respectively, ⁵⁵ while Ireland and Cyprus (the latter to a lesser extent, as explained above) are common law countries. This is an interesting observation, as it shows that the legal family of a supreme court does not necessarily determine its openness to cite the case law of foreign supreme courts.

B. ‘True comparatists’

Courts characterized as ‘true comparatists’ satisfy three criteria. First, they have made more than 10 references to foreign supreme courts. Second, they have made foreign references in more than 1% of their cases (that is, more than 10 references per 1000 cases). Third, they have made less than 66% of their citations to a single supreme court. ⁵⁶ In brief, true comparatists are therefore supreme courts that have made many references to a variety of courts. These are, as shown in Figure 3, the UK Supreme Court (58 citations), the Court of Appeal of Malta (Qorti ta’l-Appell; 299 citations) and the Supreme Court of the Netherlands (Hoge Raad; 221 citations). They shall be discussed in turn.

OPEN IN VIEWER

Figure 3.

Cross-citations of the supreme courts of Malta, the Netherlands and the UK. Explanatory note: as Figure 2.

While the absolute number of references made by the UK Supreme Court is not remarkable (58 citations), its low caseload means that it has the second highest number of references per 1000 cases out of all supreme courts (95.55). Its citations are relatively evenly distributed amongst the 14 cited supreme courts, with no apparent outliers (thus contrasting with the Irish and Cypriot supreme courts, discussed previously, that frequently cite their UK counterpart but few other courts). Perhaps ironically, the UK Supreme Court seems to have been willing to engage with ideas coming from a large spectrum of supreme courts from (as it was at the time of the data collection) other EU countries. This presents a stark contrast to the dispersion of citations of the ‘true followers’, ⁵⁷ and to smaller extent to the Dutch and Maltese supreme courts. Specifically, the UK Supreme Court has engaged most often with the German, French, Dutch, Italian and Irish supreme courts (see Figure 3). Other instances of cited supreme courts included Belgium and Austria (5 citations), Croatia (2 citations), and Finland, Greece, Lithuania, Poland, Portugal and Spain with a single citation each. ⁵⁸

There are many ways in which these results could be explained. One relates to the English conception of the judicial function. In response to questions of why they cite foreign decisions, some UK Supreme Court judges have reported that they felt a sense of duty to explain to the parties why their arguments were rejected, as well as a sense of duty to openly cite a foreign court as a matter of courtesy if they have followed the foreign court’s reasoning. ⁵⁹ In a similar vein, the House of Lords – the UK Supreme Court’s predecessor – has historically been a comparative institution, with a long tradition of consulting and referencing decisions of many other common law courts. ⁶⁰ Citing foreign decisions has been institutionally encouraged. ⁶¹ With the increased prominence of European law in the past decades, it could be that the comparative ethos ingrained into the institutional DNA of the Supreme Court was simply extended to citations of other European supreme courts. ⁶² In this European context, UK Supreme Court judges have reported that the prestige of the cited court is an important determinant of their willingness to cite, ⁶³ which also in part explains the distribution of citations we have discovered.

There are also more tangible ways in which the internationalized and Europeanized nature of the UK’s (and particularly England’s) legal order has manifested. There are many foreign, especially European lawyers working in the UK, bringing their continental background and legal thinking with them. ⁶⁴ Moreover, England is an extremely popular choice of venue jurisdiction in many international commercial contracts. ⁶⁵ Even though the jurisdiction might be English, the parties may agree on foreign, and not English law when governing the dispute. Thus, many commercial lawyers and judges in England are forced, on a regular basis, to argue cases governed by foreign law, further normalizing the presence of foreign decisions in the culture of domestic judicial decision-making and justification.

Lastly, even though UK Supreme Court judges echo the classical language restraint argument found in the comparative law literature, reporting that they feel uncomfortable consulting materials in German or French, ⁶⁶ there has long been abundant literature on foreign and comparative law published in English as the lingua franca. ⁶⁷ In this way, despite the obvious language barriers, judges have had ample opportunities to consult authoritative texts on foreign law in English. Moreover, given the relatively low caseload of the UK Supreme Court, the judges can deal with the cases it decides in more detail, including considering comparative literature and case law from other European courts.

As for the Court of Appeal of Malta, the distribution of its citations has been considerably less equal compared to that of the UK. From 2000 to 2018, the court has produced an estimated 299 citations, which translates to 31.2 citations per 1000 cases. It cited 9 different supreme courts from other European countries, with citations to the supreme courts of the UK (estimated 143 citations ⁶⁸ ) and Italy (141 citations) heavily dominating the citation pattern. Occasionally, the Court also cited the supreme courts of France (5 times), Austria (4 times), Bulgaria (2 times), as well as Belgium, Germany, Greece and the Netherlands (once).

The high number of citations and their distribution may be explained, to a significant extent, by the mixed legal nature of the Maltese legal system and its complex historical development. While its history is too rich to discuss in detail, ⁶⁹ it suffices to say that Maltese law has been affected by a variety of continental legal systems and the British common law system. The underlying continental influence was that of Roman and canon law; with more specific influences from Sicily, Italy, and France playing an important role throughout the previous millennium until British colonial rule between 1800 and 1964. ⁷⁰ The mixing of common and civil law traditions during colonial times did not make the legal system of Malta a ‘melted’ mixed legal system in which elements of the common and civil law would be interwoven. Instead, much like in Cyprus, the mix of common law and civil law was more clear-cut as they were left to operate independently in their own discrete legal areas. Private law remained in the domain of civil law, whereas common law took over public, commercial and procedural law.

It is no surprise that in a system so heavily influenced by foreign law, its jurisprudence is also ‘truly comparatist’. Indeed, this intense interplay of various influences throughout Maltese history has led ‘Maltese jurists to become natural comparatists’, ⁷¹ some have argued. One of the most significant pieces of legislation in Maltese history, the Code de Rohan (1784), included a provision permitting domestic courts to resort to ‘opinions of the Supreme and most authoritative [foreign] Courts’ when resolving controversial cases. ⁷² This comparative mindset has been carried over to today as ‘decisions of international or foreign courts applying similar legal principles’ are still considered as ‘evidential’ sources of law. ⁷³

The linguistic diversity of Malta also helps to explain the observed distribution of its citation pattern, in which Italy and the UK dominate. Maltese, a Semitic language influenced by Sicilian and Italian, has historically been the primary colloquial language in Malta, whereas Latin and later Italian were the languages of the law, even during the British rule. ⁷⁴ Despite being widely spoken, Maltese was only accepted as the official language of Malta in 1934, while Italian remained the official language of Malta until 1939. ⁷⁵ Today, Maltese and English are the official languages, with 98% of Maltese people fluent in Maltese, 88% in English, 66% in Italian and about 17% in French. ⁷⁶ Even though today judgments are drafted in Maltese, the Maltese used in courts is not the colloquial Semitic Maltese but ‘Italianate’ Maltese, heavily influenced by legal Italian. ⁷⁷ The polyglottism of the Maltese and the openness of its legal system are also demonstrated by the fact that the Maltese court refers to most of the citations of its UK and Italian counterparts in the original language.

The Dutch openness to making foreign citations is also rooted in its legal history. Already in 1999, van Erp spoke of an ‘increasingly more open legal mind and hence an ever-growing use of the comparative law method by the judiciary’ in the Netherlands. ⁷⁸ Our data support his observation. The Supreme Court of the Netherlands (Hoge Raad) produced 221 total citations, citing 15 of the other supreme courts. The highest number of citations to the German Supreme Court (114 times), followed by those of Belgium (31 times), France (21 times), the UK (20 times), Austria (13 times), Italy (6 times), Sweden, (4 times), Denmark, Greece, Luxembourg and Spain (2 times each), and the Czech Republic, Hungary, Ireland and Slovenia (once each).

Most citations were made by advocates general in their opinions, and not directly by the Court itself, which has been unwilling to refer to foreign judgments ⁷⁹ due to its argumentation style. This is based on, and closely resembles, the laconic style of the French Cour de cassation. ⁸⁰ Extrajudicially, judges themselves have openly referred to opinions of advocates general as giving more of an ‘insight into the background for the Court’s reasoning.’ ⁸¹ Indeed, the citations in the opinions are often quite thorough with citations to foreign courts in their original language (in particular those of the German, French, Belgium and UK supreme courts). Even though it is sometimes not entirely clear if a particular reference to a foreign court is decisive for the argument, such an approach reflects and confirms the general impression that Dutch judges and advocates general possess a certain linguistic prowess, enabling them to consult foreign doctrine in a variety of different languages. ⁸²

Our data supports the reports of the Hoge Raad judges that state they most often consult decisions of the German Supreme Court but also the French and UK courts. ⁸³ The data also concurs with statements made by the advocates general, who assert that they are influenced by the perceived expertise of foreign courts in specific areas of the law, referring particularly to the expertise of the German Supreme Court. ⁸⁴ Note that this is a variation of the ‘prestige of the court’ argument, akin to the situation observed for the UK Supreme Court.

Furthermore, the Dutch legal culture cannot be overlooked as an explanation of its propensity to cite foreign decisions. Dutch private law has historically been heavily influenced by foreign legal ideas and developed through comparative rationales: the old Dutch Civil Code was largely a duplicate of the French Civil Code, while the new Dutch Civil Code was developed through a rigorous comparative research of mostly Germanic civil codes. ⁸⁵ This has contributed to the development of an open legal culture and rendered it normal for Dutch courts to look at how particular issues have been dealt within the legal system from which the Dutch Civil Code originated. ⁸⁶ This historical background also helps to explain the distribution of the citation pattern. In particular it explains the citations to the German Supreme Court, but also the reasonably high number of citations of the French court (which may have been higher were it not for its formulaic and laconic argumentative style). ⁸⁷

Overall, as in the previous sub-section, this sub-section also demonstrates that comparative citations of foreign decisions are not in the exclusive domain of a single legal tradition. Fittingly, the three true comparatists (the UK Supreme Court, the Dutch Hoge Raad, and the Maltese Court of Appeal) represent the three major legal traditions found in Europe (common law, civil law and mixed law systems respectively). Our data also shows that citing foreign decisions is not unique to a single argumentation style. Here again, the three courts represent the two extremes and one in-between: we have found abundant citations in the lengthy discursive treatise of UK judges, in the moderately long decisions of the Maltese court and in the concise Dutch judgments (though the latter are supplemented by more elaborate opinions of its advocates general).

C. ‘Reluctant followers’

The following two sub-sections discuss the categories of reluctant followers and reluctant comparatists, namely courts of countries with more than 10 cross-citations, but with relative numbers of citations falling below 1%. As in the previous two sub-sections, a further distinction is made between supreme courts that mainly cite the court of one other country, and supreme courts that cite a variety of foreign courts. This sub-section discusses the supreme courts of countries with a monogenous citation pattern in more detail, given that more than 66% of their citations were to a single foreign supreme court.

Three courts are classified as ‘reluctant followers’: Belgium, Slovakia and the Czech Republic. The Supreme Court of Slovakia (Najvyšší súd) almost only referred to the Supreme Court of the Czech Republic (in 205 cases, in addition to single references to Germany and Austria); yet, it did so in a fairly large number of decisions both in absolute and in relative terms (see Figure 4). The Supreme Court of the Czech Republic (Nejvyšší soud) has, by comparison, a more heterogeneous citation style, as it referred predominantly to Slovakia (in 37 cases, that is, 66%), but also to Germany (11 times), Austria (3 times), the Netherlands (2 times) and France, Italy and Poland (once each). Likewise, the Supreme Court of Belgium (Cour de cassation, Hof van Cassatie) refers predominantly to France (in 80% of the cases), but also to the Netherlands, Germany and to some other countries in a smaller proportion.

OPEN IN VIEWER

Figure 4.

Cross-citations of the supreme courts of Belgium, the Czech Republic and Slovakia. Explanatory note: as Figure 2.

Indeed, the fact that both Slovakia and the Czech Republic fall under this category may be explained historically, given that for the most of the twentieth century (1918–1992) they formed part of a single country. Thus, they share history, legal traditions, as well as the experience of being a transition country. ⁸⁸ Prima facie, it is apparent that the Slovak court engages with Czech case law much more frequently (205 references) than the Czech court with the Slovak one (37 references). Yet, it is also worth noting that the great majority of the Slovak references to the Czech Supreme Court were made in the context of the claimant’s or the lower’s court argumentation (and, thus, one may draw a parallel to the countries in which cross-citations are mainly in the opinions of the advocates general). While a vast number of references appears in the same context, that is, the Supreme Court merely mentions the fact that one of the parties has referred to a given judgment of the Slovak Supreme Court in the prior proceedings, there are also cases where the Czech Supreme Court refers to the Slovak case law in its own argumentation. ⁸⁹ Thus, it can be said that the style of citations of the Czech to the Slovak Supreme Court is more mixed.

What may be surprising is the fact that neither the Slovak nor the Czech court refer to the Austrian Supreme Court. This is despite the fact that the Austrian Civil Code (ABGB) was effective in Czechoslovakia until 1964, and some traces of the Austrian civil law institutions remain visible to date, particularly in the Czech Republic. ⁹⁰ Moreover, the German Civil Code (BGB) has exerted significant influence on the current Czech Civil Code (introduced in 2014), which may account for the German court being the second most-cited by the Czech court. ⁹¹ In the case of Slovakia, another intriguing fact is the absence of references to Hungarian law, even though Hungarian private law was valid and applied on the territory of Slovakia until 1951. ⁹²

Overall, our findings do not confirm Bobek’s view of the Slovak Supreme Court as ‘an example of institutional and cultural settings which [is] hostile to any open use of comparative reasoning or any persuasive argument generally’, ⁹³ where ‘comparative reasoning is as good as non-existent’. ⁹⁴ While it is true that the Slovak Supreme Court does not openly engage with the Czech courts’ reasoning as an interpretative aid, it cannot go unnoticed that the parties’ and/or lower instances’ statements have at least an indirect influence.

With regards to the third country in the ‘reluctant followers’ category, the Belgian Cour de Cassation has predominantly cited its French counterpart (in 71 out of 88 cases). ⁹⁵ This is hardly surprising given that Belgian private law descends from the Napoleonian Code and is still today highly inspired by French law and case law. ⁹⁶ French legal scholars are largely referred to in Belgian scholarship, and the Belgian Cour de Cassation has often transposed developments of its French counterpart in relation to contract law. ⁹⁷ French developments and reforms were also referred to when reforming Belgian contract law. ⁹⁸ However, this tendency is one-sided since – as shown in the next sub-section – the French Cour de Cassation refers less frequently to the Belgian Court in its judgments. The other courts cited by the Belgian court are the Netherlands (9 citations), Germany (5 citations), Denmark (1 citations), the UK (1 citations) and Italy (1 citations). References to the Dutch and German court may be explained by the linguistic situation in Belgium, as Dutch and German (together with French) are the official languages of Belgium. Additionally, it may be explained by the common history that Belgium shares with the Netherlands, although Belgian law is much less influenced by German and Dutch law than by French law. ⁹⁹

To conclude, the categorization the aforementioned ‘reluctant followers’ can be plausibly explained by historical, linguistic and geographical reasons. Belgium, Slovakia and the Czech Republic are relatively small jurisdictions that have been influenced by their neighbouring countries. Much as in the case of ‘true followers’, this category encompasses countries of different legal traditions. While the influence of Slovakia and the Czech Republic was mutual, given their shared history, the relationship between the Belgian and the French court is more asymmetric.

D. ‘Reluctant comparatists’

The category of ‘reluctant comparatists’ includes the nine countries shown in Figures 5a, 5b and 5c on a different scale based on the relative frequency of their citations: Denmark, Germany and Sweden (all with more than five citations out of 1000 decisions), Lithuania, Poland, Portugal and Slovenia (all with less than five but more than 0.5 citations out of 1000 decisions) and France, Italy and Spain (all with less than 0.5 citations out of 1000 decisions). This group is highly heterogenous, comprising countries from different legal traditions and geographical regions, which makes it difficult to find a common denominator for the reasons behind their citation patterns. In what follows, we briefly present findings for each country. We first discuss Germany as the country with the highest number of foreign references, followed by countries that have been highly influenced by it (Poland, Slovenia and Lithuania). Then, we move on to the Romanistic civil law countries (Italy, France, Spain and Portugal) ¹⁰⁰ and then the two Nordic countries (Denmark and Sweden).

OPEN IN VIEWER

Figure 5a.

Cross-citations of the supreme courts of Denmark, Germany and Sweden. Figure 5b. Cross-citations of the supreme courts of Lithuania, Poland, Portugal and Slovenia. Figure 5c. Cross-citations of the supreme courts of France, Italy and Spain. Explanatory note: as Figure 2.

In Germany, the Federal Supreme Court (Bundesgerichtshof) referred to Austria in 104 out of 169 cases. Its other most frequent citations were made to the UK (19 times), Italy (12 times), France (10 times) and the Netherlands (10 times). Fewer references have been made to Spain, Belgium, Hungary, Ireland and Sweden. It is, thus, interesting to observe that with around 61.5% of citations to Austria, it is on the verge of a classification as a ‘reluctant follower’. The influence between Austrian and German court decisions is mutual (noting that the Austrian court frequently cites the German one). ¹⁰¹ However, more detailed analysis of the references of both supreme courts reveals that, while the Austrian court tends to properly employ a comparative reasoning in its judgments, the references of the German court are predominantly ornamental and do not engage with the cited judgments in detail. ¹⁰² This observation is also true for references made by the German court to other foreign courts. In general, it can be said that the German Supreme Court is often detailed in its legal argumentation and it supports its reasoning with extensive references to domestic judgments (be it to its own decisions, or to lower court instances) and academic literature. Yet, the foreign case law plays only a marginal role in its argumentation. ¹⁰³

The Slovenian Supreme Court (Vrhovno sodišče) cited Croatia in over half of its references to foreign supreme courts (35 out of 63 cases, that is, 55%). Other references included Germany (9 citations), Austria (8 citations) and Italy (8 citations). Additionally, singular references were made to the UK, the Netherlands and France. The dominance of citations to the Croatian Supreme Court may be explained by a number of factors, such as the geographical vicinity, shared history (Slovenia and Croatia were both part of Yugoslavia until 1991), and the fact that Croatian is a commonly spoken language in Slovenia. Moreover, it is interesting to note that many decisions referring to the Croatian Supreme Court ¹⁰⁴ concerned the Convention on the Law Applicable to Traffic Accidents. ¹⁰⁵ The high number of references to the German Supreme Court may reflect the influence of the German legal system, which has been traditionally considered as ‘best practice’ in the Slovenian legal academia and many pieces of Slovenian legislation have been heavily influenced by German statutes. ¹⁰⁶ The citations to the Austrian Supreme Court may be explained by the geographical proximity of Austria to Slovenia as a significant number of these cases had a cross-border element. The number of citations to the Italian Supreme Court may be due to the relatively high number of Italian speakers in Slovenia because of the big Italian community residing there. ¹⁰⁷ Overall, the study’s findings confirm that after independence, Slovenia underwent the process of ‘Europeanization’ of its legal system. ¹⁰⁸

Another ‘reluctant comparatist’ is Poland, with 24 references in total. The foreign supreme courts most frequently cited by the Polish Supreme Court (Sąd Najwyższy) were the German and Austrian Supreme Courts (9 and 7 citations) with few references to those of Greece (2 citations), France (2 citations), Italy, the Netherlands, Slovakia and the UK (1 citation each). German civil law was a model for the Polish civil code introduced in 1964 which, although with numerous changes, remains in force to date. ¹⁰⁹ The German influence is visible both in terms of general principles and concepts, as well as in the construct of specific legal instruments, and the comparison of Polish and German law is also strongly embedded in Polish academic literature. ¹¹⁰ Linguistic reasons may also explain why the German Supreme Court was the most-cited court by Polish Supreme Court judges, as German is one of the most commonly spoken foreign languages in Poland. ¹¹¹ With these considerations in mind, one might have even expected more references to the German law in the studied period. The relatively high number of citations to Austrian citations are some substantive references, appearing mainly in the context of conflict of jurisdictions, but also, interestingly, in cases concerning the interpretation of the Bills of Exchange and Promissory Notes.

The Lithuanian Supreme Court (Aukščiausiasis Teismas) made 11 cross-citations, namely to Germany (5 times), the Netherlands (3 times), Spain (2 times) and France (once). This may be interpreted as showing that the Lithuanian court has a potential interest in German law but also some other Western European jurisdictions (and it may also be noted that the relative frequency of these cross-citations is higher than that of the Polish court, see Figure 5b). Yet the information about these cross-citations needs to be qualified, as six of them are found in two near identical judgments delivered on the same day. ¹¹² Thus, it can be suggested that the foreign citation practice of the Lithuanian court is on the verge of being an ‘isolate’ (as the other two Baltic countries, discussed in the next sub-section).

As regards the Romanistic civil law countries, prior research has observed that in Italy and France cultural constraints speak against citing anything beside the relevant codes and status. ¹¹³ Indeed, the Italian Supreme Court (Corte di Cassazione) has a relatively low rate of citation with only around 2 citations out of every 10,000 decisions. ¹¹⁴ References are predominantly made to the French Supreme Court (37 citations out of 85), and then to the UK (19 citations), German (17 citations), Greek (10 citations), Austrian and Spanish courts (one each). However, many of these references were made within the context of sister judgments being delivered on the same day, meaning a chain of references was duplicated for each case. It can also be noted that half of the citations made by the Italian Supreme Court refer to the French Supreme Court, a tendency that is not mutual for the French court, as seen below.

The case law of the French Supreme Court (Cour de Cassation) has a rate of citation similar to the Italian Supreme Court with a ratio of around 2 citations per 10,000 decisions. ¹¹⁵ Since the judgments of the Cour de Cassation feature only a short deductive reasoning and rarely cite case law, ¹¹⁶ our analysis includes the opinions of the advocates general. ¹¹⁷ More specifically, out of 47 citations, 11 referred to Germany, 7 to Belgium, 6 to Spain, 5 to Italy, 5 to the UK and 4 to Ireland. Other minor references were made to the supreme courts of Bulgaria, Greece and Poland (2 citations each) and Austria, the Netherlands and Portugal (1 citation each). Hence, references to the German Supreme Court are not dominant neither in absolute nor relative terms, though a degree of mutuality may be present as the German court also referred to the French in 10 cases.

Spain is the country with the lowest relative frequency of foreign citations in the category of ‘reluctant comparatists’, ¹¹⁸ with the judgments of the Spanish Supreme Court (Tribunal Supremo) totalling only 16. Out of these, 5 were made to France, 4 to Italy, 3 to Germany, 2 to the UK, and 1 to Austria and Sweden. As a country of a largely French legal tradition, ¹¹⁹ the explanation of the French, as well as of Italian influence in Spanish private law cases is straight-forward. This citation pattern of the Spanish Supreme Court in private law cases also corresponds with the one of the Spanish Constitutional Tribunal, which referred to Italy, France and Germany most frequently. ¹²⁰

The Portuguese Supreme Court (Supremo Tribunal de Justiça) resorted to foreign supreme courts’ ruling 42 times in total, and thus with a relative frequency higher than the other three Romanistic civil law countries. ¹²¹ References were made 15 times to France, 13 times to Germany, 8 times to Spain, 3 times to Italy, twice to the UK and once to the Netherlands. This finding reflects the influence of both French and German legislative models on the Portuguese law. ¹²² While the first Portuguese Civil Code followed the model of the Napoleonic Code, German law exerted a strong impact on the further developments of Portuguese law over the course of the twentieth century. ¹²³ Furthermore, it is interesting to note that Portugal referred to Spain 8 times. By contrast, as seen above, this is not reciprocal, as no reference was made to the Portuguese jurisprudence by the Spanish Supreme Court. It is difficult to account for this difference, but one possible explanation could be that Portugal is a smaller country and therefore refers to a larger country in its judgments. Overall, some references were very laconic, made only in footnotes, ¹²⁴ while others properly engaged with the foreign judgment.

Finally, the data reveal that Denmark and Sweden did not refer frequently to one another in their case law. This is surprising given claims about a closely connected ‘Nordic legal family’, ¹²⁵ and their common history up until the 19th century. The 14 cross-citations by the Danish Supreme Court (Højesteret) are to Germany (3 citations), Sweden (3 citations), the Netherlands (3 citations), France (2 citations) and single citations to Finland, Italy and the UK. In the case of the Swedish Supreme Court (Högsta domstolen), the 16 cross-citations were split across Denmark (8 citations), Germany (3 citations), the UK (2 citations) and one each to France, Ireland and Austria. Hence, while the Nordic countries share a common history and – with the exception of Finland – a linguistic proximity, their supreme courts do not appear to form a close bond by way of citing each other. The low use of cross-citation between Nordic supreme courts might relate to the judicial style of court decisions in these countries, being relatively short and pragmatic and making little reference to academic sources. It also seems that Nordic cooperation might take other forms, such the drafting of common model laws and a shared political vision, ¹²⁶ rather than extensive judicial dialogue.

E. ‘Isolates’

We have characterized the supreme courts of eight countries as ‘isolates’. These are the courts of Croatia, Estonia, Finland, Greece, Hungary and Romania (see Figure 6), as well Bulgaria and Latvia (not displayed in this figure, as will be explained). The defining characteristic of isolate courts is the low absolute number of cross-citations of any foreign higher courts (less than 10 cross-citations). ¹²⁷

OPEN IN VIEWER

Figure 6.

Cross-citations of the supreme courts of Croatia, Estonia, Finland, Greece, Hungary and Romania. Explanatory note: following the classification scheme of Figure 1 (and different from Figures 2 to 5), this figure only refers to the absolute, not the relative, number of cross-citations.

In detail, there were 5 cross-citations by the Romanian Supreme Court (Înalta Curte de Casaţie şi Justiţie) and the Greek Supreme Court (άρειος Πάγος), 3 by the Estonian Supreme Court (Riigikohus) and two by the Finnish Supreme Court (korkein oikeus), the Hungarian Supreme Court (Kúria) and the Croatian Supreme Court (Vrhovni sud). No citations to foreign supreme courts have been identified in the case law of the Bulgarian Supreme Court (Върховен касационен съд) and the Latvian Supreme Court (Augstākā tiesa). ¹²⁸ Despite the modest number of cross-citations, isolates add a useful dimension to our discussion of geographical proximity, origins of legal tradition, judicial culture and language barriers as comparison parameters for the regional judicial dialogue between supreme courts.

The category of ‘isolates’ is geographically diverse, incorporating countries belonging to a number of European regions: the Baltics (Estonia and Latvia), the Nordics (Finland), Central Europe (Hungary) and Southeast Europe (Bulgaria, Croatia, Greece and Romania). Geographical proximity within these subgroups, however, says little about the cross-citation practices of the respective supreme courts. The supreme courts of the Baltic states do not cite each other but only cite supreme courts from Western European courts (for Estonia see Figure 6 and for Lithuania see Figure 5b in the previous sub-section). The Finnish court cited once the UK and German courts, but not any of the other Nordic courts (nor any of the Baltic ones). No cross-citations were identified between the Bulgarian and Romanian courts, with the German court again being the most frequently cited foreign court by the Romanian court (3 citations). Finally, neither Croatia nor Greece cite neighbouring foreign courts, with the Greek court again mainly citing the German court (4 citations).

As to the role of legal families, it is noteworthy that this category does not include any common law countries. However, its composition is highly heterogenous with respect to the origins of civil law tradition. At first sight, the category appears to be composed of civil law countries from the German tradition (Bulgaria, Estonia, Croatia, Greece, Hungary, Latvia). Yet, at a closer look, a significantly more nuanced picture emerges, as the countries that undertook a civil law reform in the second half of the 20th century did not strictly follow one single tradition. Examples are plenty: some Bulgarian private law statutes (notably contract law) were influenced by German law, yet others largely based on the French tradition; ¹²⁹ the Hungarian Civil Code of 1959 followed both Austrian and German traditions; ¹³⁰ the Croatian private law was in part influenced by Austrian and German (contract and property law) while other parts (principally property law) derived from Swiss and Italian law; ¹³¹ the Baltic states also not only referred to German but Swiss and Dutch civil codes when replacing the Soviet era civil codes in the 1990s. ¹³² The category, moreover, includes representatives of the French civil law tradition (Romania) ¹³³ and Nordic civil law (Finland).

Similar to the geographical subgroups, the shared civil law origins are not markedly reflected in the citation patterns. First, there are no citations of the French Supreme Court’s case law by the Romanian supreme court. The references to the case law of the German Supreme Court, albeit more frequent within the countries following the German civil law tradition, have rather a sporadic nature. As we have seen above, although the German Supreme Court is the most frequently cited foreign court by the Estonian, Greek and Hungarian courts, the Bulgarian, Croatian and Latvian courts do not cite the German Supreme Court at all. Moreover, references from the Estonian, Greek and Hungarian courts are difficult to be compared to the citation patterns within other categories due to a scarcity of such references. Hence, the well-documented cross-border influence at the legislative level does not appear to extend to the judiciary.

Finally, it is possible to analyse this category of ‘isolates’ under the perspective of former Socialist bloc countries and the question of a possible Central and Eastern European (CEE) legal tradition. The majority of the countries in this category – six out of eight ¹³⁴  – had to reform many areas of private law after the dissolution of the Eastern or Socialist bloc. The legal reform in these countries in the last decade of the twentieth century has been largely facilitated by means of cooperation with Western European states and supranational organizations. ¹³⁵ Thus, property, contract, company and commercial law reforms have used Western European civil codes, including German, Dutch, French and Swiss codes, as models to amend or entirely replace redundant laws adopted in the Communist era. ¹³⁶ Despite the well-documented sources for the new civil law provisions originating from the German, Dutch or Swiss laws, a number of considerations might prevent the supreme court judges of the ‘isolate countries’ from referring to the case law of the supreme courts of the countries from which these laws have been borrowed. Specifically, it is suggested that this limited dialogue can be framed as a set of factors defining the relative openness of legal cultures (for example, legal borrowing and cooperation), on the one hand, and the lack of such openness of judicial cultures (for example, the perception of the roles of judges), on the other hand.

The low degree of judicial openness shows, first, in the finding that there are close to no cross-citations between the supreme courts of the CEE countries discussed in this sub-section. Second, however, this only in part confirms the claim advanced by Giaro with respect to inadequacy of East-West divide in the discussion over European legal tradition. ¹³⁷ According to Giaro, ‘the neighbouring eastern peripheries never communicate horizontally with each other, but always vertically with the centre, to which they constantly aspire.’ ¹³⁸ Yet, despite some references to the German Supreme Court by the ‘isolates’, the absolute number of such cross-citations is extremely low to provide a strong support for the second part of this claim.

Third, the absence of cross-citations between the CEE ‘isolates’ could be explained by reference to the roots of national civil laws that predate the Socialist period. In spite of the introduction of new civil codes in a number of former Socialist jurisdictions, these laws take their roots from an earlier period (1918–1939) when ‘the intense circulation of western legal models throughout Eastern Europe’ took place. ¹³⁹ Hence, the shared historical conditions and similarities in the process of civil law reform in these countries do not lead to a single model of ‘CEE civil law jurisdictions’, which would otherwise prompt an enhanced dialogue between their supreme courts.

Fourth, the lack of judicial openness (despite the openness of legal culture, especially legislatively) could be further explained by other factors affecting judicial dialogue. Such factors may be the traditional roles of judges, language barriers and, possibly also, accessibility of case law. As regards the traditional roles of judges, elements of the ‘survival’ of the socialist legal tradition transposed throughout the judicial system (institutional continuity, appointment of judges and judicial argumentation) can constitute one obstacle to enhanced judicial openness. ¹⁴⁰ Language barriers constitute another obstacle, as none of the ‘isolates’ share the same language. ¹⁴¹ Finally, the low number of cross-citations may be explained by domestic publication and reporting practices and the accessibility of supreme court judgments. These potential obstacles are analysed in the following section.

A. Sources of case law

When searching for case law of supreme courts in the 28 EU Member States in the 2000–2018 period, we used a variety of types of online databases. The list of these databases was already displayed in Table 1, at the beginning of this article. ¹⁵⁰ It should be noted that this table did not aim to provide a comprehensive overview of all databases available in each Member State. Instead, it lists the database, or a combination thereof, that we have used in each state and that provides the most easy-to-search and comprehensive access to the case law in the studied period. ¹⁵¹ In addition, Table 1 included two further pieces of information: first, the originator of the database namely whether it is a public body – be it competent government ministries, the courts themselves, or any other state bodies – or a private entity; and, second, its accessibility, that is, whether the database is freely available or a subscription is needed. Table 2 summarizes these two further aspects.

OPEN IN VIEWER

Table 2.

Originators and need for subscription of databases used in this study ¹⁵² .

	Subscription not needed	Subscription needed
Public originator	Austria, Belgium, Croatia, Czech Rep., Denmark, Estonia, France (LegiFrance), France (Cour de Cassation), Germany, Greece (Areios Pagos), Hungary, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland (Sąd Najwyższy), Portugal, Romania, Slovakia, Slovenia, Spain, Sweden	[none]
Private, non-commercial originator	Cyprus, Ireland, UK	Greece (DSA Isokratis) [free for members of the Bar only]
Private, commercial originator	Finland	Bulgaria, Greece (Nomos), Italy, Poland (LEX)

As Table 2 demonstrates, an overwhelming majority of the databases we have used to collect a comprehensive sample are public and accessible free of charge. Other than the private databases we have used in Bulgaria, Greece, Finland, Ireland, Italy, Poland and the UK, all databases originate from public bodies. Additionally, other than the commercial databases in Bulgaria, Greece, Italy and Poland, which are subscription-based, all other databases we have used are available without subscription. ¹⁵³ In the cases of Bulgaria, Greece, Poland and Italy, a subscription to private databases was needed to access older case law. ¹⁵⁴ In Poland and Greece, like the UK and Ireland, private databases were preferred due to their superior search function. ¹⁵⁵ In Italy as well, a private subscription-based database was used to search the entirety of the studied period, even though jurisprudence since 2009 is freely available from a private database and case law of the last five years is available on the official website of the Corte Suprema di Cassazione. ¹⁵⁶ Thus, since 2015, all case law of supreme courts in all EU Member States can be freely accessed in public databases.

This is in line with the move in the legal information market in the past decades away from strong commercial players ¹⁵⁷ towards enabling free access to all legal materials to everyone as advocated for by various Open Law initiatives, ¹⁵⁸ the academic discourse, ¹⁵⁹ legal practitioners ¹⁶⁰ as well as by various European actors. ¹⁶¹ The EU, in particular, has provided strong financial support for the establishment of public databases of case law. Through its Horizon 2020 programme, it funds the ‘Common Portal of Case Law’, which integrates databases of most Member States’ supreme courts into a single search engine. ¹⁶² It also provided funding for the establishment of many domestic public databases, including the recently renovated Croatian, Slovenian and Maltese databases. ¹⁶³ Through this, the EU has contributed to the establishment of the European landscape of legal informatics we live in today – one that has liberated itself from the monopolies and oligopolies of private publishers of the past, towards a present and a future in which legal materials, including jurisprudence of supreme courts, are available for anyone to consult openly.

B. Completeness

In most cases, the full studied period (2000–2018) was available in the national databases, with the exceptions of Denmark (available from September 2009), Luxembourg (from 2002) and Malta (from 2001). However, many databases featured significantly fewer decisions from the early 2000s: this was the case for the databases of the Czech Republic, Luxembourg, Poland, Romania and Slovakia. ¹⁶⁴

The question is whether this is due to the lower workload of the highest courts in the early 2000s or due to the fact that there was less reporting and that fewer decisions were digitalized at the time (or a combination of both). As indicated previously, in the great majority of cases, the numbers of the decisions available in the databases corresponded roughly to the number of decisions listed on the CEPEJ database. ¹⁶⁵ Nevertheless, in many cases, it was unclear whether the national database featured all the decisions from 2000 to 2018 or if only a selection of decisions appeared. The German database states explicitly that all decisions from 1 January 2000 are available in the database. In other cases, the uncertainty is not easily resolved as neither the website of the supreme court nor the database asserts whether the database is complete or not. Additionally, no information could be found in the literature as to whether the database was complete or not (this was the case, for instance, for the Czech Republic, Hungary, Greece, Portugal and Spain). In Slovakia, the literature reveals that until the second half of 2008 the vast majority of decisions were not publicly accessible. ¹⁶⁶ In other cases, vague formulations make it hard to estimate to which extent the database is complete. The BAILII database for the Irish Supreme Court states that the database is ‘reasonably comprehensive’ from February 2001, while the Swedish one claims to contain only ‘guiding decisions’. ¹⁶⁷ Hence, we have identified not only a problem of completeness but also of transparency as to whether the database is complete or not. ¹⁶⁸ An effort for transparency was made by the Polish Supreme Court in an order of its President in 2012 stating that all decisions after 2003 must be published on the electronic database of the Court (with the exception of decisions in which the Supreme Court refused to accept the cassation complaint for consideration). ¹⁶⁹ It also seems that, in some cases, the database is updated gradually, and that recent judgments might not have all been added on the database. ¹⁷⁰

Specifically, we also examined two further aspects: (i) whether advocate general (AG) opinions were available in the examined databases (if applicable), and (ii) whether the courts issued single, consolidated judgments, and if not, whether individual opinions delivered by judges were freely and fully available in the databases. Four of the EU Member States (Belgium, France, Luxembourg and the Netherlands) have a system of advocates general that assist the decision making of the national supreme courts. ¹⁷¹ Yet, only the Dutch court made the full text of all AG opinions publicly available. In the case of Belgium, the availability of AG opinions was only partial, or rather, it was unclear whether all AG opinions were available, as in some cases these were placed in the ‘summary’ field of the final judgment. In the case of the French court, the accessibility of AG opinions was lower through the database than through a simple Google search, as the used database required a separate search for judgments and opinions. Finally, the Luxemburg court informed us that it does not make AG opinions available for confidentiality reasons (even though AG opinions were issued as per the court’s functioning). ¹⁷²

As regards the second aspect, we established that 25 of the scrutinized supreme courts issued single opinions (that is, one consolidated judgment), whereas two supreme courts published the individual judgments delivered by the judges sitting on the bench. These three cases are the UK, Ireland and Cyprus, which is in line with the general insight that common law courts are more likely to feature such individual judgments. ¹⁷³ By contrast, the Maltese supreme court which, in some respects, can also be considered to be a common law jurisdiction, ¹⁷⁴ only issues single opinions.

C. Search functions

The vast majority of the websites used for this study can be described as user-friendly and functional, allowing the identification of relevant judgments without major technical difficulties. ¹⁷⁵ That said, however, we encountered some limitations with the search functions of several search engines. For example, the website of the Dutch Supreme Court, albeit easy to navigate, did not enable the search for alternatives (in the form of ‘or’ searches). The Slovakian search engine did not allow the selection of results for both the Commercial and the Civil Law Division at the same time. In the case of the Luxembourgish Supreme Court, the search function provides only a blank search form for a word-search without any additional filters.

There were also several cases where the limitations of the search functions were of a different gravity. For example, the Hungarian website was not user-friendly, in particular with regard to the non-technologically advanced keyword search function, which was leading to false positives. The Portuguese database also turned out to be rather hard to navigate, since it was not possible to narrow down the results to private law cases and to determine the timespan. ¹⁷⁶ The greatest difficulties concerned the French website which does not provide a general search function for several chambers. Thus, the search was performed entirely through Google. In this way, while not entirely straightforward, it was possible to identify the relevant judgments of the French court and sort them out of the multiple false negatives by focusing on the files in PDF format.

As for other countries, our attempts to run the search for foreign references in the supreme court judgments through Google in order to verify and/or refine our search results led to mixed results. In some cases, for example Austria, Hungary, Italy, Luxemburg and Bulgaria, Google did not prove useful in improving the results of the search through dedicated court websites. In most cases, it simply redirected to the official database. In Estonia and Latvia, the benefits of searching through Google were limited, as only court decisions available on the webpage in summaries appeared. For some countries, for instance the Netherlands and Cyprus, the results seem under-inclusive, compared to the ones of the search engines of the courts’ portals. In the case of Ireland and the UK, a ‘Google-search’ could not be performed as the BAILII database prohibits search engines to index judgments. In the case of Belgium, due to the already mentioned access through a plug-in, searching through Google was not possible. For a number of countries, ¹⁷⁷ Google did, in principle, allow for searching the court judgments; however, it was neither more practical nor more inclusive than directly through the official websites.

As our research aimed to identify the cross-citations in matters of private law only, an important part of the inquiry was to reduce the results to private law cases. In the majority of databases, it was possible to safely divide the search into private law and non-private law cases as the search function allowed a differentiation between court divisions (or areas of law) or as the judgments themselves included such information. ¹⁷⁸ However, in some instances, the databases/judgments did not allow such a clear division. For example, in the cases of the supreme courts of the UK, Ireland and Cyprus (when using the public ‘cylaw’ database), the nature of the matter sub judice (private, criminal or other) could only be safely identified by reading through the case. In a few other instances, although the database in principle offers the possibility to limit the search results to private cases only, it was either not possible to combine this feature with other filters (for example, term-search) ¹⁷⁹ . Also, in some instances, there were reasonable doubts as to the accuracy of the results and the effectiveness of the filter. ¹⁸⁰ Thus, often, the lack of an option to reliably narrow down the search to cases dealing with private law matters meant that we verified each judgment that contained a cross-citation by hand in order to exclude non-private law cases.

Some of the problems with identifying all references to foreign supreme court judgments were also directly related to the limitations of the search functions. For example, the website of the Spanish Supreme Court produced many false positives, unless the search term was put in inverted commas to show only the exact results. This method required great precision. For example, the results from searching ‘Tribunal Supremo Federal Alemán’, did not include references to ‘tribunal federal alemán’ – both of which were valid for our data collection. On the portal of the Supreme Court of Belgium, there was no reliable option to search for a given word specifically. Filtering out the cases relevant for our study required going through many false positives (for example, results containing ‘François’ instead of ‘Français’).

Further problems did not relate to the search functions per se, but rather to the way in which judgments of a given country are typically cited. This was an issue with the references made to the UK Supreme Court (and previous decisions of the House of Lords), which was most problematic in searching for references made by the supreme courts of Ireland, Malta and Cyprus. Sometimes the citation includes ‘UKSC’ or ‘UKHL’ which will clearly indicate that the case was from the Supreme Court (UKSC) or the House of Lords (UKHL). But more often than not, a UK case will simply be cited on the basis of the law report in which it was published, for example using ‘WLR’ or ‘All.ER’, which can also cover UK judgments other than those from the UK Supreme Court (or previously the House of Lords). Thus, the only way to establish whether a cited case is a UKSC/UKHL case, or a case from any lower English court, was to manually look up every single case that is cited in a given decision. Since doing such searches for all ambiguous results was not feasible, we took a sample of 20 cases of which we thought might contain many false positives and then inferred a total number of citations from there.

More generally, another impediment to the identification of all judgments falling within the scope of our study relates to the reference style of the supreme courts. Searching for foreign citations was easier for the courts which tend to cite foreign judgments using the original names of the supreme courts and/or their common abbreviations rather than the translations. This was an effective method for instance for the judgments of the German Supreme Court. It typically uses the full foreign name of the supreme courts it referred or their abbreviation (for example, ‘OGH’ for the Austrian Supreme Court). The same is true for the Netherlands and the Czech Republic. More often, though, few results were found this way, as the courts preferred to use solely the translated names. Thus, we also searched for the translated names of the supreme courts, often using multiple variants. To give an example, in English, German Bundesgerichtshof could be found under ‘German Supreme Court’, ‘Federal Supreme Court of Germany’, ‘German Federal Tribunal’, ‘Federal Tribunal of Germany’ and so on, and the French Cour de Cassation could be referred to, among others, as the ‘French Supreme Court’, ‘French Cassation Court’, ‘French Cassation Tribunal’, ‘Cassation Court of the French Republic’, and so on.

The search for foreign court references with translated terms was even more complex in the case of some countries which have deflections (for example, Slavic languages). Accordingly, search for the references made by the supreme courts of the Czech Republic, Slovakia, Slovenia and Poland required checking the forms in nominative, genitive and accusative. This factor might have contributed to some false negatives which were not identified in our research. In addition, in a few instances the foreign name of the supreme court cited was not used in its original form, but only its translation or abbreviations of the translation. In that category, we quote indicatively, Poland and the fact that the Polish grammar has noun declension. It was therefore necessary to check the Polish names of foreign courts not only in nominative (for example, ‘Niemiecki Trybunał Federalny’ or ‘niemiecki Sąd Najwyższy’), but also in genitive (for example, niemieckiego Trubunału Federalnego or ‘niemieckiego Sądu Najwyższego’). This made this search more complex. Notably, in the case of Poland, commercial databases (for example, lex.pl) are more flexible as, even if one searches the term in nominative, the results showing this term in other cases will appear as well. By way of an example, searching for references made by the Polish Supreme Court to the Austrian Supreme Court (‘austriacki Sąd Najwyższy’) would give 1 result on the public database, and 7 results on the commercial one, for the same time period and Chamber.

Moving on to other issues that might have impacted on the reliability of the results, a reference to the search engine of the Supreme Court of Romania shall be made. During our research through the database, we had inadvertently found the judgments in which the Romanian Supreme Court refers to BGH as ‘Curtea Federală din Germania’. ¹⁸¹ But when we searched for this exact term that was used in the judgment we had found by chance (‘Curtea Federală din Germania’), the search engine provided no results. Thus, it is likely that our research on Romanian citation patterns suffers from some false negatives on account of a suboptimal search function.

Overall, the issues about the transparency and ease of access of supreme court decisions discussed in this section offer a mixed conclusion. Using various databases and search techniques, it has been possible to collect information about cross-citations from all the 28 supreme courts under investigation. Yet, for some courts, even the best sources and techniques face the limitation that the courts produce very short judgments with any possible foreign influence being left unmentioned (and thus the court not being transparent about it). ¹⁸² In addition, some limitations in terms of completeness and search functions were identified in this section. While many of those limitations could be overcome, for instance, by way of checking citations by hand, it would be helpful for both research and practice if courts were to address these limitations.