Theory and method


While constitutional reasoning can be understood in different ways, the present project uses the following definitions. Reasoning refers to the justifications that the judges publicly adduce for their decision. What makes reasoning specifically “constitutional” is that it is based on grounds that are commonly viewed as constitutional. We apply a formal definition of constitution. For this project constitution means the rule or body of rules enjoying the highest rank in a legal order (Jakab 2013).

Any attempt to analyze commonalities and discrepancies in the argumentative style practiced by constitutional judges must ultimately rest on some understanding of the motives and constraints that orient judicial decision-making (Jakab, Dyevre and Itzcovich 2017). The formalist (or legalist) model of judicial decision-making posits that judges are essentially preoccupied with the correct application of legal rules and doctrines (Schauer 1988). In this view, if judicial practices show similarities, it is because judges must apply similar rules; and if they diverge it is because the rules they are in charge of applying differ. Most constitutional comparativists do not accept the formalist account of judicial behaviour. They rather tend to discuss the role played by economic and social mechanisms as well as the influence of inherited legal structures. The approach we take in the present project draws on a realist account of judicial decision-making. According to the realist model, first laid down by Oliver Wendell Holmes (1896) judicial decision-making is determined by considerations that are not purely legal. However, we do not claim that judicial decision-making is arbitrary or that constitutional rules and doctrines do not place important constraints on judges. Instead, our project is based on the assumption that constitutional rules and doctrines do not represent the only influence on judicial conduct. Empirical research has shown that constitutional adjudication is influenced by the policy preferences of the individual judge as well as by institutional and strategic constraints (Epstein, Landes and Posner 2013; Hanretty 2012).

We conceptualize constitutional reasoning as an exercise in persuasion. We suggest that the opinion-writing practices of constitutional judges are determined by the interaction of three factors: (1) the judges’ values and policy preferences, (2) the views of the judicial role held by the judges’ audiences, and (3) the judges’ skills in crafting arguments that reflect their own preferences while remaining attuned to the expectations of their audiences.


The project involves one constitutional law scholar from each of the five Nordic countries. Each scholar is responsible for analyzing the constitutional reasoning practice in her own jurisdiction and prepare a report to serve as a basis for the comparative analysis. An Advisory Board, composed of three eminent Nordic legal scholars of the field and three other internationally renowned legal scholars with experience in applying the research methods of our project, supports the activities undertaken by the project participants.

The project studies the reasoning-practice of the supreme courts of five Nordic countries: the Swedish Högsta domstolen, the Norwegian Høgsterett, the Icelandic Hæstiréttur, the Danish Højesteret, and the Finnish Korkein oikeus. Two of the Nordic countries also have a supreme administrative court (the Swedish Högsta förvaltningsdomstolen and the Finnish Korkein hallinto-oikeus), which are also included in the analysis. Their inclusion is motivated by the fact that administrative courts also have the power to review the constitutionality of the laws they apply, thus engaging in constitutional reasoning. The inclusion of administrative supreme courts in the study also allows us to compare argumentative practices of these courts with those of the ordinary supreme courts.

In order to compare the reason-giving practices of the Nordic countries with supreme and constitutional courts in other parts of the world, the project uses the methodology of the CONREASON-project (2011-2016), adapting it to the Nordic context. That project, led by András Jakab, developed a groundbreaking conceptual and methodological framework to enable a comprehensive and systematic analysis of constitutional reasoning in 18 supreme jurisdictions (Jakab, Dyevre and Itzcovich 2017), without, however, extending its geographical reach to Northern Europe.

The project aims at a twofold comparative analysis: (1) among Nordic countries, (2) between Nordic countries and other countries in the world. In particular, the project investigates, among others, the following issues:

  • Do Nordic supreme courts put more or less emphasis on certain key concepts?
  • How often do they use non-legal arguments?
  • How often do they rely on precedents?
  • How often do they refer to certain types of sources - such as preparatory works, foreign legal materials and academic writings?
  • How often do Nordic judges dissent from their peers?
  • Do they use some special key concepts specific to Nordic constitutionalism?

A quantitative approach is needed to identify and compare argumentation patterns, while a qualitative analysis aims to ensure maximum comparability across the jurisdictions.

Since there is a large number of cases in the long history of all Nordic supreme courts, a comprehensive analysis of all constitutional decisions is not feasible. For this reason, we have to make a selection. We document the reasoning practices of a court on the basis of its leading cases, i.e. cases where the rulings are deemed the most important in the legal community, with no time limit as to the year of the delivery of the judgments. Leading (or landmark) cases tend to set the tone of a court’s jurisprudence, as they often provide the lens through which court watchers recognize the defining traits of a court’s approach to constitutional argumentation.

For the selection of the leading cases we employ a method often used in social sciences: expert opinion. Thus, the selection is made by the project participant responsible for the report on the given jurisdiction. We assume that a relative consensus usually exists as to what decisions constitute leading judgments. To enhance the list’s accuracy, each researcher is also required to designate five constitutional law scholars of their own jurisdiction to review her choice of cases.

As regards the qualitative element of the research, every project participant responsible for the analysis of one Nordic country fills in a detailed questionnaire. The questions cover not only the court’s style of reasoning but also its institutional configuration and broader political environment, which will inform the comparison. The questions directly addressing the court’s argumentation style fall in three broad categories:

  • the dominant topical and argumentative structure of constitutional opinions,
  • the sources of constitutional law and the repertoire of interpretative methods,
  • key concepts and generic constitutional doctrines.

As to the quantitative element, the dataset will consist of a summary of the argumentative characteristics of 280 landmark constitutional decisions of the seven examined courts. In particular, we aim to estimate the incidence of the examined opinion characteristics in all five Nordic countries. We use cluster analysis to identify possible subgroups, which allows us to see if the reasoning practices of certain Nordic supreme courts are more similar to each other than to other Nordic supreme courts. We will also carry out cluster analysis on an aggregated dataset of this project and the parent project’s dataset to discover differences and commonalities of Nordic constitutionalism with other constitutional traditions of the world.

Finally, an examination of the temporal distribution of cases and of argumentative practices will enable us to draw conclusions about the historical development of constitutional reasoning in Northern Europe, and to assess the extent to which these fit into the global trends.