Comparing Courts across Countries with Docket Type

Empirical research on comparative judicial politics has seen tremendous growth in recent years (see, e.g., Kapiszewski and Taylor Reference Kapiszewski and Taylor2008; Case Reference Case2012). However, whereas the majority of this empirical research focuses on single-country studies, the number of empirical studies that encompass two or more countries is markedly less prevalent (e.g., Cooter and Ginsburg Reference Cooter1996; Hönnige Reference Hönnige2009; Wetstein et al. Reference Wetstein, Ostberg, Songer and Johnson2009; Weiden Reference Weiden2011; Hanretty Reference Hanretty2012; Bricker Reference Bricker2017; Weinshall, Sommer, and Ritov Reference Weinshall, Sommer and Ritov2017).

Of those studies examining judicial behavior across multiple countries, scholars have shown a particular interest in analyzing attitudinal factors. Yet, institutional constraints can be an equally powerful catalyst for moving judicial behavior. Comparative scholarship focusing on the institutional effects of docket control on behavior is much more limited (Fontana Reference Fontana and Dixon2011). Previous scholars have leveraged the comparative advantage of American state governments (Eisenberg and Miller Reference Eisenberg and Miller2009; Brace et al. Reference Brace, Yates and Boyea2012) or changes in docket rules within a single country (Eisenberg and Huang Reference Eisenberg and Kuo-Chang2012) to model the effects of docket type. Nevertheless, as Fontana (Reference Fontana and Dixon2011, 626) notes, “scholars have not studied comprehensively the different docket control structures in place around the world.” Our study adds to this literature by analyzing these institutional effects across two countries: Denmark and Norway.

Existing scholarship on the relationship between docket type and judicial behavior suggests that this institutional feature affects dissent and reversal rates, conditions the impact of litigant status, and influences the magnitude of attitudinal behavior on courts. The following section presents this literature and generates predictions about docket type and judicial behavior.

A Most Similar Systems Design: Danish and Norwegian Supreme Courts

The Danish and Norwegian Supreme Courts offer a favorable and instructive institutional comparison of the impact of docket type on judicial behavior in the European context. Importantly, these two courts are situated in two countries that share several key political, cultural, legal, and institutional similarities (see table 1). Norway and Denmark are geographically close to one another and have populations of roughly equal size. Both countries are parliamentary democracies that have been characterized by political scientists as consensual democracies (Knudsen and Rothstein Reference Knudsen and Rothstein1994; Lijphart Reference Lijphart1999). Indeed, Norway and Denmark, together with Sweden, Iceland, and Finland, are often considered to be a distinct group of countries in political science studies (designated the “Nordic model”), characterized by generous welfare states funded by high taxes, high interpersonal trust, universal social rights, and equality (see, e.g., Knutsen Reference Knutsen2017). Their legal systems are so similar historically that the Danish ambassador to the United States in 1950 “pointed out [that] the laws of the Scandinavian states are closer to each other than are the laws of the [different] states of the United States” (Orfield Reference Orfield1952, 773).

Table 1.

Institutional Detail: Danish and Norwegian Supreme Courts

Note.—Rows 1–7, 11, and 18 are gathered from Sunde (Reference Sunde2017). ECHR = European Convention on Human Rights; EU = European Union; EEA = European Economic Area.

From 1537 to 1814, Norway was largely subject to Danish rule (the Danish-Norwegian union). During this period, Norway and Denmark had a common Danish-Norwegian Supreme Court, which was based in Copenhagen. When the union was dissolved in 1814, this common arrangement had to be terminated, and the Norwegian Constitution of 1814 established the Norwegian Supreme Court as an independent court of last resort (sec. 88 of the Norwegian Constitution). The newly established court was, however, modeled according to its Danish counterpart, with a similar organization, working method, and oral process (Schei Reference Schei2015).

The two courts are still very similar today. Both courts are a part of what scholars refer to as the Nordic legal system—a civil law system exhibiting common-law features (Lindblom Reference Lindblom2000; Hirschl Reference Hirschl2011; Sunde Reference Sunde2015). Most interestingly, although the Nordic legal system considers precedents to have an important influence on the law, precedents are not legally binding, as is the case in Anglo-American common law (Blume Reference Blume2011; Hirschl Reference Hirschl2011). On their own initiative, the Norwegian and Danish Supreme Courts have redefined their legal and social task so that their stated tasks now match their de facto law-making functions. In 1994, the Norwegian and Danish Supreme Courts defined their tasks as creating “legal unity, legal clarification and legal development” (Sunde Reference Sunde2017, 72–73).

Whereas the Danish court consists of 19 justices (one chief justice and 18 associate justices), the Norwegian court consists of 20 justices (one chief justice and 19 associate justices). Both courts have general jurisdiction over all cases involving criminal, civil, administrative, and constitutional law. The Danish Supreme Court is located at the top of a three-tiered court hierarchy consisting of 24 district courts, three courts of appeals, and one supreme court. The Norwegian Supreme Court is located at the top of a three-tiered court hierarchy consisting of 63 district courts, six courts of appeals, and one supreme court. The recruitment of justices to the two supreme courts follows similar institutional procedures (Christensen Reference Christensen, Magid, Melchior and Stocholm2011; Grendstad, Shaffer, and Waltenburg Reference Grendstad, Shaffer and Waltenburg2015; Sunde Reference Sunde2015). In both countries, independent judicial appointment boards announce vacancies, handle applications, and make a final recommendation to the government, who then forwards the recommendation for a formal appointment by the king in council. Both appointment boards recommend justices based on professional criteria. To date, the respective governments have always appointed the justices recommended by the judicial appointment boards.

Once an appeal is granted, the decisional environments in the two courts are similar, except for how the courts treat seniority and assign opinion writing. Cases in both courts are decided by majority vote in two rotating panels consisting of five justices. Panels are created in a semirandom (or arbitrary) fashion. In both courts, the role of presiding justice is given to the chief justice or, in the absence of a chief justice, the most senior justice (Pagter Kristensen Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011; Sunde Reference Sunde2015, chap. 7). Some differences exist between the two courts, however. In the Norwegian court, the presiding justice is assigned the responsibility of presenting case facts and legal sources for the other justices, whereas in the Danish court, this task is assigned to the opinion writer. Additionally, in the Danish court, the initial task of writing the opinion is given to the most junior justice. By contrast, in Norway, opinion authorship follows a schedule such that all justices on the panel, except the presiding justice, take turns writing the decision of the court. Moreover, the style of the final judgments is somewhat different; whereas judgments of the Danish court appear more impersonal and are written as the opinion of the full court, judgments of the Norwegian court are written in a more personal style (similar to judgments of the US Supreme Court). When it comes to voting on a case, it is the least senior justice in Denmark who initiates the voting process, while in Norway this task is given to the most senior justice. Thus, seniority plays a different role in the Norwegian and Danish courts. Finally, in Denmark, cases raising more principled legal questions are decided in enlarged panels, most often composed of seven justices. In Norway, cases raising more principled legal questions concerning international law or constitutional issues are decided in the grand chamber with 11 justices or en banc with a maximum odd number of justices.

The similarities in the Danish and Norwegian legal systems also relate specifically to tax law. Scholars often refer to a Nordic model of taxation (Olsen Reference Olsen1992; Ganghof Reference Ganghof2006; Livingston Reference Livingston2016) and classify Denmark and Norway in the same tax law family (“Northern European Nordic”; Thuronyi, Brooks, and Kolozs Reference Thuronyi and Brooks2016, chap. 2). The Nordic model of taxation, writes Livingston (Reference Livingston2016, 78), “involves relatively high and only modestly progressive individual income tax; significantly lower corporate dividend and capital gain taxes; and, increasingly, environmental taxes that have a pollution-reducing as well as a revenue-raising function … and a high degree of sophistication in tax administration.” As such, for purposes of our analysis, the population of tax cases accessing the two courts every year should be quite similar.

Key Institutional Difference between the Two Courts: Docket Control

Whereas similarities between Denmark and Norway dominate, there is one crucial difference: the Norwegian court has near-complete discretionary control over its case docket, while its Danish counterpart exercises little discretionary control. The different docket categories on the Danish and Norwegian Supreme Courts are illustrated in figure 1. The figure shows four different case trajectories, which we classify according to four categories: (1) complete mandatory docket, (2) mandatory docket of principled cases, (3) discretionary docket by external appeals boards, and (4) discretionary docket by justice panels.

Figure 1.

Four docket categories in tax cases on the Norwegian and Danish Supreme Courts. The box indicates the court of original jurisdiction. (Sources: Pagter Kristensen [Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011] and Sunde [Reference Sunde2017].)

The Norwegian Supreme Court belongs to the “discretionary docket by justice panels” category. Since 1996, the court has had near-complete discretion over its case selection. The selection is administered by the appeals committee, which screens all appeals. The committee is composed of three justices who rotate among all of the justices. An appeal is granted if only one of the three justices finds that the case promises “importance beyond” itself (Bårdsen Reference Bårdsen2018). In describing the cases selected by the Appeals Committee, Norwegian Supreme Court Justice Arnfinn Bårdsen (Reference Bårdsen2018, 11) stated: “The need for a clarifying ruling from the Supreme Court will typically emerge because the legal material leaves no definite answer regarding what the law is—it is open to debate whether the applicable norm says A or B; professionally trained lawyers can, on justified grounds, argue for both—or even more—positions.”

In contrast, the Danish Supreme Court has never had complete discretion based on justice panels. Instead, at various points in time, the court has found itself in one of the other three categories. Before a procedural change occurred in 2007, some cases on the Danish court followed the “complete mandatory docket” category. Under this configuration, all cases involving government departments in Denmark started in the second-instance court (the appeals court) and would automatically be heard by the supreme court if one of the parties appealed the case.

After the 2007 procedural change, most of the cases in the Danish court belonged to the “mandatory docket of principled cases” category. Under this configuration, cases would start in the district court. However, the district court could transfer the case to the appeals court if they believed that the case raised a “principled” legal question. Therefore, many tax cases would still start in the appeals court, and consequently, litigants would have a right to appeal the case to the supreme court. Another reform occurred in 2014, after which the appeals court was granted authority to decide whether a case should start in the district court or in the appeals court. That is, the final discretionary power to define a case as “principled” now lay at the court of appeals. Nevertheless, in the case of a loss at the appeals court, the parties would still have a right to appeal the case to the supreme court.

In addition, some cases in Denmark arrive at the supreme court through an external appeals board (hence, these cases belong to the “discretionary docket by external appeals boards” category). The external appeals board (the Procesbevillingsnævnet) has discretionary power similar to the Norwegian appeals committee and is only able to grant an appeal if a case is of a “principled character.” The board consists of five members: one supreme court justice (who does not serve on the supreme court while on the board), a district court judge, an appeals court judge, a noted lawyer, and a law professor.Footnote ¹ The decision to grant appeals is made by simple majority (Nørgaard Reference Nørgaard2001). The external appeals board is not the main source of cases for the Danish Supreme Court. Rather, it functions as the mechanism that identifies principled cases that lower-court judges failed to identify (Pagter Kristensen Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011, 143–44).

Although the recent institutional changes do not grant discretionary power to the Danish justices in a similar way as on the Norwegian court, the intention has clearly been for the Danish court to concentrate its resources on the most principled cases (Pagter Kristensen Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011, 137). Despite this intention, however, we can expect that case-selection competence and incentives for identifying principled cases vary among lower-court judges and external bodies. The variation in case-selection competence and incentives should, in turn, influence the different proportion of principled cases between the dockets of the courts. Hence, the number of dissents and reversals, as hypothesized in H1 and H2, will also show variation between the two courts.

The differences in selection procedures should also affect the magnitude of attitudinal behavior on the two supreme courts. To test whether docket type influences relative party status (H3) and the expression of preferences (H4), we address the question of government deference. In both countries, a long-standing debate exists concerning whether the supreme courts are too deferential toward the government. Deference can stem from an overly insular recruitment of justices from government administrative offices (Kjønstad Reference Kjønstad1997; Gøtze Reference Gøtze2010) or as a consequence of the governments’ superior resources and experiences relative to private litigants (Ryssdal Reference Ryssdal2007).

The recruitment of justices to the Danish and Norwegian courts is generally perceived as nonpartisan (Sunde Reference Sunde2015; but see Grendstad, Shaffer, and Waltenburg Reference Grendstad, Shaffer, Skiple and Waltenburg2015), but scholars have suggested that justices’ preferences toward the government party vary according to their preappointment occupational socialization (Gøtze Reference Gøtze2010; Skiple Reference Skiple2015; Skiple et al. Reference Skiple, Grendstad, Shaffer and Waltenburg2016). Empirical studies of cases concerning public economic interests versus private economic rights on the Norwegian Supreme Court have shown that justices with experience from the legislation department of the executive branch are more likely to favor the government party (Skiple Reference Skiple2015), whereas academics are more likely to favor the private party (Grendstad, Shaffer, and Waltenburg Reference Grendstad, Shaffer and Waltenburg2015, chap. 5; Skiple et al. Reference Skiple, Grendstad, Shaffer and Waltenburg2016). Zimmer (Reference Zimmer2015) found similar patterns in a descriptive analysis of 124 income tax cases from 1997 to 2012. Accordingly, if docket control indeed affects the degree to which the judges express their personal preferences, this should be evident from the magnitude to which the justices’ preappointment legal socialization influences their attitudes under the various docket regimes.

Data, Measurement, and Procedures

The data collected by the authors in this study encompass all civil law tax cases before the Danish court from 2006 to June 2017 and the Norwegian Supreme Court from 1996 until 2016. The data for the Danish court build on and extend the Danish Supreme Court Database (McKenzie, Bentsen, and Skiple Reference McKenzie, Bentsen and Skiple2016), while the data on the Norwegian Supreme Court build on the Norwegian Supreme Court Database (Grendstad, Shaffer, and Waltenburg Reference Grendstad and Shaffer2016). For this study, the data on the Norwegian Supreme Court comprise 207 court decisions, and the data on the Danish Supreme Court comprise 330 court decisions. For the Norwegian Supreme Court, we use an existing variable measuring whether a public party is up against a private party in a tax case (Bjørnebekk Reference Bjørnebekk2015; Grendstad et al. Reference Grendstad and Shaffer2016).Footnote ² In the Danish data, tax cases were identified by the Danish tax authorities being a party to the case.

Apart from the fact that the two countries belong to the same tax law family, which ensures compatibility of analysis across the two courts, there are three other reasons why we study tax cases. First, studying tax cases allows us to include the “complete mandatory docket” category in our data because tax cases in the Danish court arrived at the court through complete mandatory dockets until the 2007 reform and beyond (Pagter Kristensen Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011, 140–41). Second, tax cases usually pit individuals or corporations against the government, allowing us to test H3 (mandatory dockets increase the chance of success for high-status litigants, compared to discretionary dockets). Third, tax cases in the Scandinavian courts reason well with the phenomenon of government deference, indicating that we can make reasonable assumptions about justices’ preferences toward government deference based on earlier studies (e.g., Grendstad, Shaffer, and Waltenburg Reference Grendstad, Shaffer and Waltenburg2015).

We extend our data on the Norwegian court back to 1996 for two reasons. First, because the Norwegian court handles fewer tax cases than the Danish court, a longer time span ensures a sufficient sample size. Second, 1996 was the first full year of tax cases after the comprehensive 1995 reform in Norwegian criminal procedures. Since 1996, the supreme court has exercised full discretion over the selection of criminal cases. Although the 1995 reform did not have a direct influence on the handling of civil tax cases, it was a game-changing reform for the court moving toward being an active developer of the law (Sunde Reference Sunde2015). Thus, by limiting the sample to cases decided in the period after 1995, we secure a higher degree of internal measurement validity.

Dependent Variables: Nonunanimity, Reversal, and Government Win

We include four different dependent variables in our analyses: nonunanimous decision, dissent, reversal, and government win (see table 2 for descriptive statistics for both dependent and independent variables).

Table 2.Descriptive Statistics for Dependent (DV) and Independent (IV) Variables

To measure nonunanimous decisions, we rely on a dichotomous variable that takes the value 1 if the decision contains any form of disagreement between justices and the value 0 if the decision is unanimous. In the construction of the nonunanimous decision variable, we note the difference between dissents (disagreement about the substantive outcome of the case) and concurrences (disagreement about the rationales leading to the outcomes of decisions). In Norway, justices frequently offer concurrences (Bentsen Reference Bentsen2018). This practice is not common in Denmark (Pagter Kristensen Reference Pagter Kristensen, Per Magidp, Stocholm and Tamm2011, 173). Hence, we develop two different variables for nonunanimous decisions on the Norwegian court: one variable that includes all forms of disagreement (dissent and concurrence) and one variable that only includes dissents.

To measure reversal, we include a dichotomous variable that takes the value 1 if a majority of the justices on the court vote to reverse the lower-court decision, indicating that the appellant wins the case. The variable is coded 0 if there is no reversal.

To measure government wins, we include a dichotomous variable that takes the value 1 if the public party wins the case and the value 0 if the private party wins the case.Footnote ³ To test H4 (attitudinal effects), we code the data at the level of justices. In a nonunanimous case, the variable takes the value 1 if a justice votes in favor of the government party and the value 0 if a justice votes in favor of the private party.