Author(s): Michael Hein
Source: European Journal of Legal Studies, Volume Issue ()


How Do Types of Procedure Affect the Degree of Politicization of European Constitutional Courts?
A Comparative Study of Germany, Bulgaria, and Portugal*


Michael Hein and Stefan Ewert**


Do different types of procedure affect the degree of politicization of constitutional courts in European democracies? We argue that they do and we find evidence that supports this assumption in our analysis of the German, Bulgarian and Portuguese courts. We show that two features of the types of procedure lead to a higher politicization of court decisions: lower legal requirements on the part of the applicants and broader opportunities for them to weaken political opponents. This kind of moderating effect appears equally for all groups of applicants.


Keywords:constitutional courts, constitutional review, politicization, types of procedure


I. Introduction


Do the types of procedure affect the politicization of constitutional courts? We argue that they do and we find evidence that supports this assumption with regard to courts in three European democracies: Germany, Bulgaria and Portugal. These cases show that the lower the legal requirements on the part of the applicants, and the greater the opportunities for attacking political opponents within a particular type of procedure, the higher the level of politicization of the constitutional court decisions is. For quite some time, political scientists investigating constitutional courts have focused on political influences on constitutional review. These influences result from the position of constitutional courts on the interface between law and politics.[1]Moreover, they are likely to occur, since constitutional law is highly open to interpretation. In particular, when political actors handle interest-based conflicts before the constitutional court, political influences are virtually inevitable. However, political influences threaten the adjudicatory nature of constitutional adjudication.[2]This may cause problems of legitimacy, especially if these influences call the judges' independence – and hence the rule of law – into question.

Against this background, many studies have analyzed the politicization of constitutional courts. By politicization we mean that a constitutional court decision is not – or not exclusively – taken based on legal criteria. Instead, it is (co‑)determined by political influences, especially the judges' political party affiliations and their policy preferences. Moreover, these preferences can mirror ethical or religious beliefs or the judges' socio-cultural backgrounds.[3]The causes of politicization can be manifold: direct impacts on judges by political actors, the appointment of judges, the strategies and interests of the parties involved in constitutional disputes (including the judges themselves), and the judges' and the courts' understanding of their roles.


It follows that political and legal culture, the content of the contested legal norms, public attention, and types of procedure exert moderating effects in the sense that these variables do not themselves cause politicization but modify its degree.[4]These moderating factors thus 'channel' the political influences on constitutional adjudication.


However, previous political science analyses have mainly focused on politicization with regard to the appointment of judges,[5]the strategies and interests of the parties involved,[6]and political actors' attempts to directly control the courts.[7]By contrast, studies of other factors are largely missing. This also holds true for types of procedure. They can be understood as regulating the access to constitutional courts, the possible contents of constitutional review, and the consequences of the court decisions.[8]They thus define who may bring what kind of legal problems under which circumstances before the constitutional court, and which judicial consequences a judgment will have.Consequently, the types of procedure substantially configure the constitutional court's case law. Since politicization may impair the adjudicatory nature of constitutional review, the concrete configuration of the types of procedure is of vital importance for the legitimacy of a constitutional court. Procedures that potentially facilitate politicization play a rather problematic role. In contrast, procedures that potentially mitigate politicization would be preferable from a rule of law perspective. Therefore, substantial knowledge on how the types of procedure influence politicization is not only desirable for the scientific analysis of constitutional courts, but also for political and legal practitioners who deal with the establishment or reform of constitutional courts.


Nevertheless, political science has very rarely taken different types of procedure into account so far, at least with regard to courts following the European model of concentrated judicial review. Silvia von Steinsdorff analyzes the types of procedure concerning the political consequences of court decisions in eight post-socialist constitutional courts in Central and Eastern Europe.[9]Steven Schäller examines the impact that types of procedure have on the occurrence of prejudices in the decisions of the German Federal Constitutional Court.[10]Apart from these examples, however, the relevant literature provides only a few scattered arguments on the impact certain types of procedure have,[11]an overview of 'ancillary powers' of constitutional courts,[12]and a descriptive typology of types of procedure.[13]Aside from those, many comparative political science studies focus only on selected types of procedure, most commonly abstract review procedures or disputes between state bodies.[14]

Therefore, this article examines whether and in what manner the different types of procedure influence the politicization of constitutional court decisions. As outlined above, we consider the types of procedure as a moderating variable that channels politicization: we assume that the level of politicization varies among different types of procedure. We conduct our analysis using a Most Different Systems Design. From all concentrated constitutional courts in European democracies as of 2010, we selected the German Bundesverfassungsgericht (Federal Constitutional Court), the Bulgarian Konstitucionnen săd (Constitutional Court) and the Portuguese Tribunal Constitucional (Constitutional Tribunal). We analyzed all decisions of these constitutional courts from 1991 to 2010 for Germany and Bulgaria, and from 2005 to 2010 for Portugal. On the basis of two theory-driven hypotheses, we divided all types of procedure into four groups, which we expect to have different levels of politicization.


In the following, we first present our concept of politicization and discuss the challenges of distinguishing between politics and law (II.). We then proceed to develop our two hypotheses and describe the different types of procedure that can be found across European constitutional courts (III.). We then operationalize politicization (IV.) and explain our case selection (V.). Subsequently, we illustrate the empirical data and present our results (VI.). Finally, we interpret our results in the light of our hypotheses, discuss the validity of our findings beyond the countries analyzed, and provide an outlook on further research (VII.).


II. The Concept of Politicization


'Politicization' is one of the most frequently used terms in scientific research on constitutional adjudication. It can be found in numerous studies that deal with political influences on constitutional courts and supreme courts (and also on lower-level ordinary courts) in practically all regions of the world.[15]Due to the tradition of conceptualizing judges as 'politicians in robes',[16]i.e. as politically interested actors that do not systematically differ from other political actors, the concept played a rather minor role in US American political science for quite a long time. However, even there the concept of politicization has gained in importance in recent years.[17]


In general, politicization describes the process of 'making previously non-political persons or issues political', i.e. 'the expansion of politics, especially of the power to take socially binding decisions and to penetrate previously non-political fields such as private life or private economic activity'.[18]
We assume that constitutional adjudication is basically a legal and not
a political process. However, it is subject to political influences for at least two reasons. First, constitutions regularly contain politically controversial issues such as the normative foundations of the constitutional order or the basic rules of the political process. In particular, when political actors handle interest-based conflicts before a court, political influences are virtually inevitable.


Second, constitutional law is often quite general in its formulation and therefore highly open to interpretation. Admittedly, this openness differs depending on the kind of constitutional norm. For instance, rules are generally considered less disputable than standards – and these in turn are both deemed less disputable than principles.[19]Nevertheless, even if 'provisions are formulated as clearly and as coherently as possible, they can raise questions when it comes to solving a concrete case.'[20]Therefore, constitutional judges frequently have a broad scope for decision-making and they do not necessarily need to base their decisions solely on legal criteria. Quite often, constitutional courts are even required to decide on matters for which they cannot find clear answers in the constitutional text. As Robert A. Dahl put it with regard to the US Supreme Court, 'it is an essential characteristic of the institution that from time to time its members decide cases where legal criteria are not in any realistic sense adequate to the task […]; that is, the setting of the case is "political".'[21]Therefore, the judges may be (and sometimes inevitably are) politically influenced, most notably by individual political party affiliations and policy preferences. These influences can also reflect ethical and religious ties, ideological beliefs, and the socio-cultural backgrounds of the judges.


In sum, politicization of constitutional adjudication means that a court decision is not or not exclusively made on the basis of legal criteria, but is (co‑)determined by political influences. Politicization thus refers to the instances in which judges either are (consciously or unconsciously) influenced in their search for the most convincing legal solution to the constitutional matter before them by their political preferences or considerations regarding political appropriateness, or even decide a case on the basis of political criteria and then prepare a legal reasoning to support this.[22]


However, distinguishing between political and legal criteria is the main challenge in the study of politicization – both theoretically and empirically.[23]Theoretically, even precise definitions as provided e.g. by sociological systems theory[24]are highly disputed. Empirically, it is impossible to identify political influences in single court decisions since the real motives of the judges cannot be discovered unequivocally. These difficulties notwithstanding, it seems reasonable to assume that if political influences did not play any role in the work of a court, there should be no long-term correlations between any causes of politicization and the court's adjudication. Under a number of restrictions, it is therefore possible to approach politicization by observing certain characteristics of constitutional court decisions such as the probability of success of oppositional claims[25]or the distribution of dissenting votes.[26]Since the former would be restricted to a few types of procedure only, we have opted for the latter (see section IV.).


III. Types of Procedure as 'Channels of Politicization': Two Hypotheses


As mentioned in the introduction, the types of procedure regulate the access to constitutional courts, the possible contents of constitutional review, and the consequences of the court decisions. Accessmay be given to state bodies, natural and legal persons, organizations (e.g. political parties), and the constitutional court ex officio. The contents of constitutional adjudication are manifold (see table 1 below). The consequences of the decisions can be classified into three legal categories: non-binding, binding only for the parties involved in the proceedings (inter partes), or generally binding for the constitutional order in total (erga omnes). Additionally, the implementation of a decision in practice may differ substantially from the legal provisions, seeing as constitutional courts do not have their own executive capacities and are dependent on other state bodies' willingness to comply.


1. Hypothesis 1: Legal Requirements

Our first hypothesis refers mainly to the access to and the content of constitutional adjudication. It takes the legal requirements that the different types of procedure impose on the applicants into account. The hypothesis follows an argument that has been present in the literature for a long time, but has until now not been proven empirically. It goes back to Alexis de Tocqueville's seminal study on 'Democracy in America'. Tocqueville develops his argument based on the two characteristics of judicial power: the obligation to decide on individual cases and the requirement of a claim:


As long, therefore, as the law is uncontested, the judicial authority is not called upon to discuss it, and it may exist without being perceived. […] If a judge in deciding a particular point destroys a general principle, by passing a judgement which tends to reject all the inferences from that principle, and consequently to annul it, he remains within the ordinary limits of his functions. But if he directly attacks a general principle without having a particular case in view, he leaves the circle in which all nations have agreed to confine his authority, […] he ceases to be a representative of the judicial power.[27]


Also with regard to the US Supreme Court, John Stuart Mill reaffirmed this argument. Referring directly to Tocqueville, Mill stated that the Supreme Court


decides only as much of the question at a time as is required by the case before it, and its decision, instead of being volunteered for political purposes, is drawn from it by the duty which it can not refuse to fulfill, of dispensing justice impartially between adverse litigants.[28]


Transferred to modern constitutional adjudication, this means that types of procedure, such as the individual constitutional complaint or the concrete review, in most cases remain in the legal field since they are bound
to individual legal disputes. Hence, these types of procedure should lead to
a rather low level of politicization. By contrast, types of procedure such as the constitutional interpretation or the abstract review more likely protrude into the political sphere due to their lack of case specificity and thus function as channels for a comparatively high level of politicization.


Recent research on constitutional courts has repeatedly been reconsidering this argument. As Nuno Garoupa and Tom Ginsburg put it: 'Whereas concrete review 'judicializes' constitutional courts, preventive review has the opposite effect. Preventive review makes a constitutional court
less judicial and more political'.[29]This argument can be generalized as follows: the likelihood that the decision-making process will be politicized depends on whether or not the applicant has to assert a violation of, or a threat to, their constitutional rights or competencies. If this is the case, the proceedings can only be opened in order to decide a concrete judicial dispute. In that context, the claimants' opportunities to pursue political interests are rather limited. Consequently, the judges' discretion is also clearly defined and limited. Thus, our first hypothesis regarding the legal requirements of the different types of procedure reads as follows:


(H1)If a type of procedure does not require the claimant to assert a violation of or a threat to their constitutional rights or competencies, the politicization level will be higher than if the claimant does have to make such an assertion.


2. Hypothesis 2: Political Opportunities

Our second hypothesis considers the political opportunities of different types of procedure. Thus, it refers primarily to the consequences of
the court's decision. Tocqueville already pointed out that constitutional judges are 'invested with immense political power'. Their power to examine the constitutionality of a law 'gives rise to immense political influence'.[30]This influence can be exerted in line with or in contradiction to the interests of political actors. In contrast, those actors might try to use constitutional court proceedings for political goals, i.e. for purposes other than those intended by the constitution. Instead of the repeal of an unconstitutional situation, the settling of a legal dispute, or the protection of the constitutional order from its enemies, claimants might go to court primarily in order to reduce their political opponents' powers, to unseat political opponents, or even to eliminate them as political opponents.


We assume that the different types of procedures channel these political intentions – the causes of politicization – in different ways. Types of procedure that are commonly considered 'political' seem to be particularly suitable for these purposes, such as the removal from office, impeachment procedures, the party ban, or the dispute between state bodies. For instance, the failed impeachment proceedings against the previous Romanian President Traian Băsescu (2007 and 2012) showed a remarkably high degree of politicization.[31]In particular, this became evident when the parliament initiated these procedures. The voting behavior of the deputies and senators ran almost exactly along party lines, even though on paper the procedure is not a political recall, as it deals with infringements of the constitution. Similarly, the presidential impeachment proceedings conducted in US history – in particular those of Andrew Johnson (1868) and Bill Clinton (1998/1999)[32]– were heavily dominated by the conflicting political interests of the two 'blocs' in the Congress.[33]Since in both countries only the parliament can initiate an impeachment procedure, it provides political actors with an opportunity to struggle for power by legal means, even though those means were designed to protect the constitution.[34]In general, we assume for all above-mentioned 'political' types of procedure that there is a high probability of political interests being slipped in the cases and that this cannot be easily neutralized by the judges.[35]


In contrast, types of procedure such as the constitutional interpretation,
the actio popularis, the concrete review, or the individual constitutional complaint seem to be more 'judicial'. In these types of procedure, political opponents cannot directly be threatened in their position by the judgment. Therefore, it is much less likely that claimants try to transfer their political controversies to the constitutional court. Admittedly, decisions in these types of procedure can have indirect political consequences for political opponents (see the next subsection). Nevertheless, 'individual judges might be prone to political reasoning in a dispute between state bodies, while concrete review proceedings might evoke more legal reasoning.'[36]


Again, we can generalize this argument for all types of procedure: Regarding the probability of a politicized decision by the constitutional court, the types of procedure have a moderating effect according to the opportunities they provide for the claimants to directly weaken the position of their political opponents. Thus, our second hypothesis regarding the political opportunities of the types of procedure reads as follows:


(H2)If a type of procedure provides the opportunity to weaken political opponents directly, the politicization level will be higher than if such an opportunity is not provided.


3. The Expected Degrees of Politicization of the Different Types of Procedure

Combining both hypotheses, we can cross-tabulate all types of procedure requiring a claim.[37]The types of procedure in field (a) meet both hypothesized conditions. Accordingly, we expect the highest level of politicization in this group. The fields (b) and (c) meet one of the conditions each. Thus, we expect a middle-range degree of politicization in these two groups. However, while hypothesis 1 defines an absolute difference, hypothesis 2 defines only a relative distinction: The requirement to assert
a violation of, or a threat to, one's constitutional rights or competencies (H1) exists or does not. Yet, while the possibility to weaken political opponents directly also exists only in certain types of procedure, it is almost always possible to weaken political opponents in an indirect fashion. This is particularly the case for the abstract review and the constitutional interpretation. These types of procedure are not only usable to decide conflicts of jurisdiction. In fact, the political opposition might restrict the legislative scope of the parliamentary majority by means of a successful lawsuit, thereby weakening the governing actors.


In sum, we expect hypothesis 1 to have a stronger impact on politicization than hypothesis 2. Consequently, we assume a higher degree of politicization in group (b) than in group (c). Finally, the types of procedure in group (d) meet none of the conditions favoring politicization. Therefore, we expect the lowest degree of politicization here. Taken as a whole, we expect that the degrees of politicization will rank from high to low as follows: (a) →(b) →(c) →(d). Across the European constitutional courts we can find 15 different types of procedure requiring a claim.[38]They are presented in table 1 and will be described in the following more closely.

Table 1: Expected degrees of politicization through types of procedure



Legal Requirements

no violation of/threat to applicant's constitutional rights or competencies necessary

violation of/threat to applicant's constitutional rights or competencies necessary

Political Opportunities

direct weakening of political opponents possible

a: high

·         ban or non-registration of a political party

·         ban or non-registration of an organization

·         removal from office/ impeachment/ withdrawal of a parliamentary mandate due to offenses against criminal or constitutional law

·         appeals against the financial audit of political parties or election campaigns

·         forfeiture of basic rights

c: medium-low

·         dispute between state bodies

·         scrutiny of elections/withdrawal of a parliamentary mandate due to offenses against the election law

·         appeals against elections to or deliberations of the governing bodies of political parties

direct weakening of political opponents impossible

b: medium-high

·         constitutional interpretation

·         abstract review

·         actio popularis

·         (quasi-actio popularis)

d: low

·         concrete review

·         individual constitutional complaint

·         (quasi-actio popularis)

Source: Authors' own compilation.

a.Applicant's Constitutional Rightsor Competencies Not Affected/Direct Weakening of Political Opponents Possible:

We expect the highest degree of politicization in group (a). This group consists of procedures where the claimants do not have to assert a violation of, or a threat to, their own constitutional rights. However, these types of procedure are very appropriate as tools in political disputes. The threat to apply them alone can be used to weaken political opponents. Moreover, instead of clear statutory violations, constitutional judges mostly only have to identify an abstract 'unconstitutional behavior'. Thus, a decision based on judicial criteria alone is often impossible. Instead, the judges are left with a large scope for interpretation that is susceptible to political influences.[39]A special case in this group is the uncommon type of procedure 'financial audit of political parties and election campaigns'. This type of procedure can be found in Portugal, where the constitutional court exofficio reviews the finances of parties and election campaigns, and the parties involved can appeal against these decisions. The latter procedure is part of our empirical analysis, since our hypotheses are applicable here.


b.Applicant's Constitutional Rightsor Competencies Not Affected/Direct Weakening of Political Opponents Impossible:

We expect a medium-high degree of politicization for the constitutional interpretation, the abstract review, the actio popularis, and – under specific conditions – what some scholars call the 'quasi-actio popularis'.[40]The constitutional interpretation is a peculiarity of some post-socialist constitutional courts. In this type of procedure, state authorities are entitled to request the interpretation of any constitutional norm without any specific prerequisite. On the one hand, several post-socialist constitutions introduced this type of procedure with regard to the legal uncertainty during the transition to democracy and the rule of law. On the other hand, its potential politicization contradicts the intention of this type of procedure, because no concrete legal dispute is required. Additionally, the court can be forced to work legislatively and to declare one possible elaboration of a constitutional norm as the only possible option.


The abstract review is a common type of procedure among almost all concentrated constitutional courts. Apart from obligatory (ex officio) abstract review, there are two forms requiring a claim: preventive (a priori) and repressive (a posteriori) review. In both cases, only state bodies can initiate the procedure. The court then has to decide on the constitutionality of an act (be it an ordinance, a law or even a constitutional amendment) or an international treaty. We expect a medium-high degree of politicization since the claimant's constitutional rights or competencies do not have to be affected. Instead of an application of legal norms to concrete circumstances, the court is forced to review the challenged provision in an abstract manner. Moreover, typically the parliamentary opposition (or affiliated actors) bring politically controversial acts before the court. Hence, as the former German constitutional judge Helmut Simon pointed out, it is


sometimes difficult to see, who actually needs help. The parties primarily reiterate their positions from the legislative process and continue their dispute before the court. The court easily takes on the role of arbitrator between opposing value judgments and prognoses, without being better equipped or qualified for this purpose than the parliament.[41]


Government and opposition simply continue their controversy before the constitutional court. Thus, it seems obvious that the claimants in such abstract review procedures are mainly politically motivated.[42]


The actio popularis is a special type of abstract review where everybody (or at least every citizen of the respective country) can assert the unconstitutionality of a legal norm without being affected personally. The so-called 'quasi-actio popularis' differs from the genuine actio popularis insofar as this type of procedure imposes certain legal requirements on the applicant, in particular a substantiated 'legal interest'.



c.Applicant's Constitutional Rightsor Competencies Affected/Direct Weakening of Political Opponents Possible:

This group consists of three rather different types of procedure: the dispute between state bodies, the scrutiny of elections including the withdrawal of a parliamentary mandate due to offenses against the election law, and appeals against elections or deliberations of the governing bodies of political parties. The last mentioned type of procedure only applies to the Portuguese case. We expect a medium-low degree of politicization in this group, but lower than in group (b). On the one hand, claimants can initiate these procedures only if they assert a violation of, or threat to, their constitutional rights or competencies. On the other hand, the proceedings can be used primarily to negatively affect political opponents, e.g. to weaken other political parties or rivals within one's own party.[43]Even if the constitutional judges have distinct judicial criteria for their decision in such cases, they often become entangled in these political controversies and can hardly escape their dynamics.


d.Applicant's Constitutional Rightsor Competencies Affected/Direct Weakening of Political Opponents Impossible:

Finally, we expect the lowest degree of politicization for the types of procedure that meet none of the two conditions favoring politicization. This applies to the concrete review, the individual constitutional complaint and – under certain circumstances – the 'quasi-actio popularis'. Depending on its particular configuration, the latter procedure could belong to group (d) instead of group (b) (see above). This is the case if the legal requirements are shaped in such a manner that the applicant (at least de facto) has to be affected in their constitutional rightsor competencies. Thus, the judges have to decide particular judicial disputes in these types of procedure. Therefore, they are


triggered by the application of the disputed legal provisions to concrete circumstances and thus require the claimant to have been impaired in his own rights. Such a review, which can take into account the effect and scope of the provisions in legal reality, most closely corresponds to the function of a court.[44]


Generally, it can be assumed that judicial interests motivate the claimants in these types of procedure,[45]not least since 'political' state bodies usually do not have the right of petition here. Furthermore, the constitutional court decides within a clearly defined legal frame due to the case specificity.


IV. The Operationalization of Politicization


As outlined in section II, we use the judges' published dissenting opinions to analyze political influences on the court empirically. The majority of European constitutional courts enable the judges to publish their individual dissenting votes and/or the voting result nowadays.[46]Hence, each judge is allowed to take a position that differs from the judges' majority publicly visible.[47]Of course, the reasons for such a divergence can be manifold. Nevertheless, a political motivation or (unconscious) orientation is a likely reason for many cases:


The practice of allowing dissent in constitutional courts signals that they are not entirely judicial but at least partly political in character. Allowing separate opinions and dispensing with unanimous voting sends a costly signal, because it means the ideologically important goal of having a single correct answer is being sacrificed. Consensus in a constitutional court thus expresses 'legal' decision-making rather than 'political' decision-making.[48]


Even if dissent naturally appears in all collegiate courts, the majority of European legal systems permit their publication only for constitutional courts.[49]Some states such as France or Italy even prohibit the publication of dissenting opinions for their constitutional judges. In these countries,


practitioners have argued that dissenting opinions, by demonstrating that judicial outcomes are not automatic, create the perception that judicial decisions are motivated by […] political considerations rather than […] legal or jurisprudential considerations, thereby undermining judicial independence.[50]


Thus, dissenting opinions can indicate the partly political character of constitutional adjudication. For our measurement, we calculate the dissenting ratio (DR) as the share of publicly dissenting judges among all judges taking part in the decision. The DR theoretically ranges from 0.0 to 1.0. Cases with a DR > 0.5 may occur when an absolute or qualified majority of all judges is required or when more than half of the participating judges express dissenting opinions, but related to different aspects of the decision.[51]We calculate the DR mean for each type of procedure and for each of the groups (a) to (d) in order to be able to describe differences in politicization. Using a relative measurement prevents a bias that might result from the absence of certain judges. In addition, it makes a comparison of different constitutional courts and especially the identification of consensus-orientated and conflict-orientated courts possible. However, a direct comparison of the DR values of different courts is not valid in our study design, as we cannot consider all different causes and moderators of politicization. Nevertheless, the direction of the effect of the types of procedure should be similar in very different settings. Thus, we provide an indirect comparison: We compare the ranks of the types of procedure and of the different groups concerning the dissenting ratio. If, for example, the concrete review procedures in all countries have a distinctly lower DR than the abstract review procedures, this would clearly indicate a systematic relation – i.e. a higher degree of politicization of the latter type of procedure.


However, as Chris Hanretty has recently shown for the Estonian Supreme Court, 'judicial disagreement need not to be political'.[52]Instead, there is one crucial empirical precondition and one methodological limitation for using the dissenting opinions in order to analyze the moderating effects of the types of procedure. Empirically, the validity of the indicator DR is given only in cases for which it correlates with the main causes of politicization. With regard to the appointment of judges, previous studies have shown that in Germany,[53]Bulgaria,[54]and Portugal[55]the decision-making behavior of the judges indeed correlates with their individual party affiliations and the positions of the political parties that appointed them.[56]


To test the validity of our indicator, we furthermore examined the relation between different applicants and the dissenting ratio. Concerning the strategies and interests of the parties involved in constitutional disputes, one would expect a high degree of politicization with parliamentary claims because in most cases they are lodged by the opposition following (or expecting) political defeats against the governing majority. In contrast, lawsuits initiated by ordinary courts, prosecutors, or natural and legal persons should have a rather low level of politicization, since these actors normally do not primarily pursue political goals when going to the constitutional court. In the cases of Germany, Bulgaria, and Portugal the dissenting ratios fulfill the outlined expectations. According to our dataset, constitutional court decisions initiated by the parliaments show noticeably higher dissenting ratios than judgments stemming from ordinary courts, prosecutors, or natural and legal persons (see table 2).[57]In sum, we can assume the outlined empirical precondition for using the dissenting ratio as measurement for politicization to be fulfilled in the three countries under investigation.


Table 2: Comparison of the dissenting ratios of the different types of claimants







0.1002 [46]

0.171 [181]

0.215 [13]



0.147 [62]

0.050 [453]


0.0157 [333]

0.130 [23]


natural and legal persons

0.0002 [96,093]


0.014 [3,137]


Source: Authors' own compilation. Means of the claimants, number of cases per claimant [in square brackets], standard deviation (in brackets). N.e. = not eligible to apply.


The second condition is a methodological limitation. As mentioned in section II, the dissenting ratio – or any other measuring instrument one might use – cannot measure the extent of politicization in individual cases, but only reflect the differences in the degrees of politicization in long-term observations. In a single constitutional court decision, politicization can never be proven beyond doubt. It is simply not possible to discover the judges' real motives (whatever the contents of the judgment and the dissenting opinions might be). Even in cases where politicization seems obvious, the possibility remains that the distribution of votes only accidentally complies with the expected degree of politicization according to the composition of the bench.[58]


In contrast, the court's caseload as such does not restrict the validity of the indicator with regard to our research question. Of course, a high caseload leads ceteris paribus to a lower percentage of published dissenting opinions (see below). The more cases judges have to deal with, the less time they have to formulate dissenting opinions. Yet, there is no reason to assume that the caseload has an impact on the distribution of the dissenting opinions across the different types of procedure. Instead, an equal distribution would just prove the null hypothesis, that is, show that the types of procedure do not modify the politicization of constitutional court decisions.


Nevertheless, there might be a distorting effect on our measurement stemming from 'follow-up' or 'photocopy' judgments. If a court has once decided on a key matter (e.g. the interpretation of a certain constitutional provision), all cases coming before that court relating to this matter later on will be decided in line with the initial judgment (as long as the court does not change its view).[59]If the precedent decision was accompanied by dissenting opinions, and the dissenting judges also take part in the follow-up judgments, two (mutually non-exclusive) possibilities exist: Either these judges repeat their dissents, or they 'give up' and bow to the majority's opinion. Either of those possibilities might distort the measurement of politicization. If dissent is repeated in a different type of procedure, one might say that this repeated dissent should not be counted since in this unique constellation, the latter type of procedure is obviously not moderating the degree of politicization. If judges keep their dissenting opinion but do not repeat it publicly, while the type of procedure remains the same, one might contrarily say that this covert dissent should be counted because causes and moderators of politicization recur as well.


However, the information necessary for coding the data as outlined is either not accessible at all or only available by means of a thorough content analysis of all court decisions. Nevertheless, we can control for the possible distortion of the follow-up judgments on the dissenting behavior by selecting cases with different caseloads. A problematic effect of the follow-up judgments can only be relevant in courts with high caseload since in case of low caseload, the number of such judgments will also be quite low. If we find our hypothesized dissent pattern according to the types of procedure in cases with low, medium, and high caseloads, we can therefore reasonably interpret the differences in the dissenting ratio per type of procedure as differences in the degree of politicization.


V. Research Design and Case Selection


In order to answer our research question we use a Most Different Systems Design. As is common in this kind of research, we use simple comparisons of the means to analyze the relationship between the types of procedure as moderating variables and politicization.[60]Our complete[61]survey of all court decisions during the 20 (Bulgaria and Germany) and six years (Portugal) under investigation allows for valid empirical conclusions on the relationship of types of procedure and the probability of dissenting opinions.


We selected our cases from the universe of a total of 19 concentrated constitutional courts in democratic territorial states in Europe as of 2010.[62]The period under our investigation is 1991 to 2010, in order to be able to include post-socialist countries and their specific paths of transition. We excluded hybrid regimes and autocracies, since the judges' autonomy of decision has to be regarded as limited in these states. Finally, we included only courts that allow the publication of dissenting opinions during the whole period under investigation. Following the logic of the Most Different Systems Design, we selected three courts according to the following three criteria:


1.       The courts should represent different European regions in order to vary possible historical, political-cultural and legal-cultural influences.

2.      The countries should be of different ages concerning democracy, the rule of law, and the constitutional courts, since we expect a strong influence of these ages both on the judges' and courts' understanding of their roles, and the contents of the claims.

3.       The caseload should vary distinctly in order to validate the operationalization of politicization and to confirm our assumption that even courts with high caseloads show an unequal DR distribution among the different types of procedure.


Additionally, the selected courts should represent as many different types of procedure as possible, at least one of each group (a) to (d).


As a result of applying these criteria, we selected the German Federal Constitutional Court, the Bulgarian Constitutional Court, and the Portuguese Constitutional Tribunal for our study. The German court has a very high caseload and is one of the oldest constitutional courts among the consolidated democracies of Central Europe. Bulgaria represents the case most different from Germanywith a young court in a consolidating post-socialist democracy in Eastern Europe with an exceptionally low caseload. Very different from these two cases according to all above-mentioned selection criteria is Portugal, which represents another group of constitutional courts: The Constitutional Tribunal has a median caseload and is a middle-aged court in a consolidated, but younger democracy in Southern Europe. However, we had to reduce the period under investigation in this case to the years 2005–2010. Portugal reformed the appointment of judges and several types of procedure in 1998 and again in 2004/05.[63]Therefore, aggregating and comparing the DR before 2004 is impossible. This is especially true for the financial audit of political parties and election campaigns since the rules of this type of procedure were changed substantially. Nevertheless, the number of cases in the period shortened to the six years from 2005 to 2010 (altogether 3,858 decisions) still allows for the inclusion of the Portuguese case in our study.


Table 3 lists the types of procedure common to the three constitutional courts under investigation. However, two of these types of procedure have not been applied in the analyzed period (ban or non-registration of an organization, and removal from office/impeachment/withdrawal of a parliamentary mandate due to offenses against criminal or constitutional law). Hence, we analyzed ten different types of procedure from all the four groups (a) to (d). With regard to the concrete review, some explanatory remarks seem necessary here since this type of procedure is shaped quite differently in the three countries: While in Germany every ordinary court can refer any question of constitutionality of a certain legal norm to the constitutional court, only the two highest ordinary courts are allowed to do so in Bulgaria. In Portugal, traditionally every court can adjudicate a legal norm as unconstitutional, but only with inter partes effect for the case at hand. After the rejection of a complaint against such a decision, the parties to the dispute as well as the prosecution office can raise an objection before the Constitutional Tribunal. Hence, unlike in the other two countries, concrete review procedures take the lion's share in Portugal (2005–2010: 3.412 cases/93.2 percent). However, the first-instance judgments of the Constitutional Tribunal also run inter partes only. Only if the Tribunal declared a norm unconstitutional in at least three cases, the prosecution office or a constitutional judge can initiate an abstract review in order to reject it with erga omnes effect.[64]These procedures thus somewhat mix elements from abstract and concrete review. Therefore, although we coded them as concrete review procedures in a first step, we consider them in detail later on (see below).


Table 3: Types of procedure at the constitutional courts in Germany, Bulgaria (both 1991–2010) and Portugal (2005–2010)


Types of procedure








ban or non-registration of a political party




ban or non-registration of an organization




removal from office/impeachment/withdrawal of a parliamentary mandate due to offenses against criminal or constitutional law




appeals against the financial audit of political parties or election campaigns




forfeiture of basic rights





constitutional interpretation




abstract review




actio popularis




(quasi-actio popularis)





dispute between state bodies




scrutiny of elections/withdrawal of a parliamentary mandate due to offenses against the election law




appeals against elections or deliberations of the governing bodies of political parties





concrete review




individual constitutional complaint




(quasi-actio popularis)




Source: Authors' own compilation. Types of procedure in brackets were not applied during the period under investigation.

VI. Data and Results

1. Caseload and Dissenting Ratio

As mentioned in the previous section, the caseload of the three courts varies substantially. While the German Federal Constitutional Court took 96,853 decisions in 1991–2010 (mean 4,843 per year), of which 96,614 are included in our analysis, the Bulgarian Constitutional Court issued only 297 judgments in the same period (15 on average per year), of which 289 are usable for testing our hypotheses. The Portuguese Constitutional Tribunal holds a median position with 3,858 decisions in the period under investigation (2005–2010, mean 643 per year). We included 3,659 of these decisions in our analysis.[65]As we assumed, the caseload is inversely proportional to the share of published dissenting opinions. In Germany, the dissenting ratio is only 0.00032 for the whole period of analysis, but 0.15838 in Bulgaria and – again on a median position – 0.02159 in Portugal. However, there is no correlation between caseload and dissenting ratio over time, i.e., in all three countries, increasing numbers of decisions do not lead to decreasing dissenting ratios (and vice versa).[66]


Even if these figures do not reflect the different reasons for politicization and the contextual (moderating) factors, they seem to confirm the general knowledge of the respective courts. The German Federal Constitutional Court is generally acknowledged as highly consensus-oriented,[67]while the Bulgarian Constitutional Court is widely seen as characterized by strong and public conflicts.[68]The Portuguese Constitutional Tribunal appears again as a median case. Although it features a party political polarization, it has developed a noticeable consensus-orientation among the judges since the mid-1990s.[69]


2. The Degrees of Politicization of Different Types of Procedure

Table 4 shows the means of the four groups of the three constitutional courts' types of procedure. Even if the number of observations varies substantially between the different types of procedure and the countries (in square brackets in table 4)[70], the dissenting ratios – interpreted as an indicator for different degrees of politicization – differ as hypothesized in our theoretical assumptions in the cases of Germany and Bulgaria.  Procedures of group (a), which allow for a direct weakening of political opponents and do not oblige the applicant to assert a violation of or threat to their constitutional rights or competencies, feature the highest degree of politicization in both countries. Types of procedure of groups (b) and (c), which fulfill only one condition of our hypotheses each, have a medium DR, with distinctly higher scores for group (b). The types of procedure of group (d), which do not allow for a direct weakening of political opponents and oblige the applicant to assert
a violation of or a threat to their constitutional rights or competencies, feature the lowest degree of politicization.


Table 4: Comparison of the dissenting ratios of the different groups of types of procedure






Group a        
applicant's constitutional rights or competencies not affected / direct weakening of political opponents possible

0.245 [7]

0.398 [2]

0.000 [19]

Group b
applicant's constitutional rights or competencies not affected / direct weakening of political opponents impossible

0.080 [47]

0.163 [284]

0.190 [59]

Group c
applicant's constitutional rights or competencies affected / direct weakening of political opponents

0.015 [202]

0.097 [13]

0.025 [169]

Group d
applicant's constitutional rights or competencies affected / direct weakening of political opponents impossible

0.0003 [96,358]

0.075 [9]

0.018 [3,412]

Source: Authors' own compilation. Means of the four groups, number of cases per group [in square brackets], standard deviation (in brackets).


In the case of the PortugueseConstitutional Tribunal, we also see a declining DR from group (b) to (c) and (d). However, group (a) differs from the other two countries and our theoretical assumptions. In this group, the overall dissenting ratio (and the ratio of every single decision in this group) is DR=0.0. Apart from this exception, though, the comparison of the means of the four groups in the three countries confirms a moderating influence of the types of procedure in accordance with our expectations. Since this pattern is by and large observable in all three cases with their broad differences in caseload, it can be interpreted as showing differences in the degree of politicization.


In the following, we lookinto the three cases and the different types of procedure separately. This detailed analysis offers additional information to our overall results. In the case of the German court, two types of procedure do not fit our theoretical assumptions (see figure 1). The forfeiture of basic rights (in group a) and the scrutiny of elections (group c) feature the lowest possible degree of politicization with a DR of 0.0. With regard to the former type of procedure, however, only one decision was taken in the period under investigation. The court rejected an application of the federal government against two extreme right-wing politicians unanimously with the plausible (and genuine judicial) argument,


that prison sentences, to which both defendants were convicted in the context of their right-wing extremist activities, have been suspended after the application. After considering all circumstances, the criminal courts have predicted that the defendants will no longer militantly pursue their right-wing extremist convictions.[71]


Considering the singularity of this case, however, it cannot be interpreted
as an empirical rebuttal of our hypotheses.

Figure 1: Germany: Aggregated dissenting ratio per type of procedure



Source: Authors' own collection. In brackets: number of decisions observed.


All 97 scrutiny decisions of the German Federal Constitutional Courtbetween 1991 and 2010 were also taken without any public dissent. This unexpected result can partly be explained by the rejection of many complaints against election results for an obvious formal reason: they were not supported by at least 100 registered voters, as was required by the Law on the Federal Constitutional Court (§ 48, para. 1) up until 2012.[72]Still, it remains remarkable that among these 97 unanimous decisions a number of highly politically controversial cases can be found, such as the decisions on the negative voting weight, which forced the parliament to reform the election law,[73]or the decision on the use of electronic voting machines.[74]


In the case of the Bulgarian Constitutional Court, all types of procedure confirm our hypotheses without exception (see figure 2). The two party ban procedures during the period of investigation were highly politicized.
The mean of DR=0.398 is the highest aggregated dissenting ratio of all types of procedure in the three countries.[75]In 1992, a minority of five against six judges refused a party ban application against the Movement for Rights and Liberties, a partywhich primarily represents the Turkish minority in Bulgaria. In 2000, the Constitutional Court banned the splinter party OMO Ilinden-PIRIN, which represents the Macedonian minority in Bulgaria, with a 9-to-3 vote. Both litigations were clearly politically motivated.[76]


Figure 2: Bulgaria: Aggregated dissenting ratio per type of procedure



Source: Authors' own collection. In brackets: number of procedures observed.


With regard to the disputes between state bodies, it might be surprising that only two decisions were taken in this type of procedure during the whole period under investigation. In contrast, there were no less than 66 decisions in constitutional interpretation procedures. This can be explained by the high requirements for the former type of procedure. Therefore, applicants prefer – whenever it is possible – to ask for a general constitutional interpretation, although they de facto aim at resolving a concrete conflict.[77]Thus, these cases are often 'hidden' disputes between state bodies.


Contrary to the German and Bulgarian cases, no party ban procedure took place before the Portuguese Constitutional Tribunal in the period under investigation (see figure 3). However, there were 19 decisions on appeals against the financial audit of political parties or election campaigns
– the most frequent type of procedure in group (a) in all three countries. Against our expectations, all of these decisions were taken unanimously. Two arguments might explain these findings. First, contrary to all other group (a) types of procedure, no political opponents face each other as appellant and defendant before the court in this case. Instead, the affected political actor faces the constitutional court itself. Furthermore, the court isobligated to start the financial audit ex officio, before the negatively affected parties or candidates may raise objections. Thus, political strategies and interests of opposing parties as a reason for politicization are lacking to a large extent.

Figure 3: Portugal: Aggregated dissenting ratio per type of procedure



Source: Authors' own collection. In brackets: number of procedures observed.


Second, the judges do not have to interpret fairly open constitutional norms. They rather have to review whether clear-cut provisions of the financing of the Political Parties Act has been observed. Unlike in the case of the party ban procedure, the judges' role is thus similar to the one of judges in ordinary courts, who have relatively little scope for interpretation. This second argument might also explain the fact that the Constitutional Tribunal took all decisions on appeals against elections or deliberations of the governing bodies of political parties (group [c]) unanimously, even if the number of observations is rather low (seven decisions).


The other three types of procedure that appeared in the Portuguese case confirm our hypotheses. As outlined in the previous section, there is a special case of abstract review procedures: these proceedings could be initiated by the prosecution office or a constitutional judge when a legal norm was declared unconstitutional inter partes in at least three concrete reviews in order to achieve a general rejection of this particular norm. In the period under our investigation, 20 such proceedings took place. As mentioned above, we coded these cases as concrete reviews in a first step since they are based on concrete disputes. If we analyze the cases as a separate type of procedure, however, the dissenting ratio (DR=0.098) is clearly higher than in the rest of the concrete reviews (DR=0.018/3,392 cases), but distinctly lower than for the abstract reviews (DR=0.190). This result reflects the fact that this type of procedure combines elements of concrete and abstract review. It also indicates the usefulness of dissenting opinions for analyzing politicization.


When we compare the three cases on the level of the individual types of procedure, two more results stand out. First, in all three countries the abstract reviews (group [b]) are politicized to a much higher degree than all types of procedure of group (c), i.e. disputes between state bodies, scrutiny of elections, and appeals against elections or deliberations of the governing bodies of political parties. This result even persists if we include the Bulgarian constitutional interpretation procedure as 'hidden' disputes between state bodies into group (c). This confirms our assumption of a systematically stronger influence of the legal requirements hypothesis (H1) compared to the political opportunities hypothesis (H2). Second, we can confirm the prominent assumption outlined above that political motives and influences play a much stronger role in abstract than in concrete reviews. The dissenting ratios of abstract reviews are distinctly higher than those of concrete reviews in all three countries.


3. Do the Types of Procedure Only Mirror the Claimants' Strategies and Interests?

Before we conclude our results, we would like to discuss one obvious objection against our findings. One could argue that the different dissenting ratios for the types of procedure only mirror different strategies of the claimants. On the one hand, this effect could result from the fact that certain types of procedure can de jure only be initiated by certain claimants. Additionally, the appellants could de facto use the different types of procedure to a varying extent. Thus, our findings would only confirm the moderating effect of the types of procedure concerning the different strategies and interests of the parties involved.



Table 5: Ranking of dissenting ratios per type of procedure and appellant



Rank 1

Rank 2

Rank 3

Rank 4







Dispute between state bodies
0.0096 [13]

Const. complaint
0.0002 [96,025]

Scrutiny of elections
0.0000 [55]



Party ban
0.5710 [1]

Abstract review
0.1406 [8]

Dispute between state bodies
0.079 [37]


Federal Government

Party ban
0.2855 [2]

Dispute between state bodies
0.0000 [2]

Forfeiture of basic rights
0.0000 [1]


State Government

Party ban
0.1903 [3]

Abstract review
0.0750 [35]

Dispute between state bodies
0.000 [10]








Party ban
0.398 [2]  

Abstract review
0.191 [134]

Const. Interpret.
0.104 [37]

Scrutiny of elections
0.095 [8]


Const. Interpret.
0.079 [8]

Abstract review
0.000 [1]

Dispute between state bodies
0.000 [1]



Abstract review
0.180 [15]  

Const. Interpret.
0.108 [7]




Abstract review
0.181 [11]

Const. Interpret.
0.111 [3]

Concrete review
0.075 [9]


Prosecution Office

Abstract review
0.152 [49]

Const. Interpret.
0.135 [10]

Scrutiny of elections
0.114 [3]







Individuals/Organizations/Party members

Scrutiny of elections
0.025 [161]

Concrete review
0.013 [2,950]

Financial audit of parties and campaigns
0.000 [19]

Appeals against elections in political parties
0.000 [7]


Scrutiny of elections
0.077 [1]

Concrete review
0.067 [9]



Source: Authors' own collection. Comparison of the means. In square brackets: number of procedures observed.


However, we argue that the types of procedure also have a moderating effect on the politicization of the constitutional adjudication that equally affects all appellants. In order to prove this assumption, we control the influence of the appellants by analyzing the rank order of the degrees of politicization of each type of procedure for the different appellants. If these rank orders correspond with the rank orders of all appellants (see Figures 1‑3), this would confirm a moderating effect of the types of procedure on the contents of the claims and the judges' work independent from the strategies and interests of the claimants. This additional test is possible for all groups of claimants that have access to (and actually used) at least two different types of procedure. This is in fact the case for eleven actors in the three countries during the period under our investigation.


Even if the number of observations per claimant is rather low in a couple of cases, the results clearly confirm that the effect of the types of procedure is independent from the claimants (see table 5). In ten of the eleven cases, the rank order of the dissenting ratios corresponds exactly with the overall picture for the respective country. Only in the case of the Bulgarian government, it does not. Thus, our data do not only show that the claimants influence the degree of politicization of constitutional jurisdiction by their strategic use of certain types of procedure. They also demonstrate that the types of procedure modify the claims and the following court decisions for all appellants in the same way.


VII. Conclusion and Outlook


Constitutional adjudication can be politicized by manifold factors, such as direct impacts on judges by political actors, the appointment of judges, the strategies and interests of the parties involved in constitutional disputes, and the judges' and the courts' understanding of their roles. Against this background, we have conceptualized the types of procedure as a moderating variable 'channeling' those political influences. Indeed, our analysis has shown that the types of procedure have a substantial moderating effect on the politicization of constitutional court decisions. Although there have been manifold theoretical and methodological challenges in distinguishing between law and politics, and bearing in mind the preconditions and restrictions outlined in section IV, our operationalization of politicization by means of the dissenting ratio, i.e. the share of publicly dissenting judges among all judges taking part in the decision, allows us to draw this conclusion. Thus, our theoretical hypotheses can largely be considered confirmed: the lower the legal requirements of a type of procedure are, and the more opportunities to attack political opponents a type of procedure provides, the higher is ceteris paribus the politicization of the respective judicial decision. Additionally, the absence of the legal requirement to assert a violation of, or a threat to, the claimant's constitutional rights or competencies has a stronger impact on politicization than the direct possibilities to weaken the political opponents. Thus, types of procedure such as the concrete review, the individual constitutional complaint, or the dispute between state bodies seem to be preferable to procedures like the abstract review, the constitutional interpretation, or the party ban when it comes to the legitimacy of constitutional courts.


Since we used a Most Different Systems Design, it furthermore seems reasonable to expect similar politicization patterns at other concentrated constitutional courts in democratic states in Europe as well. The analysis of three very different courts from this universe of cases enabled us to prove our theoretical assumptions on a solid empirical basis. All the three cases show similar results. To test our findings in a broader setting, future research should integrate other world regions, other regime types and other periods. Additionally, qualitative case studies seem to be reasonable in order to analyze why four individual types of procedure do not confirm our hypotheses (i.e. the scrutiny of elections, and the forfeiture of basic rights in Germany; the appeals against elections and deliberations in political parties, and appeals against the financial audit of political parties or election campaigns in Portugal).


More generally, future scholarship should look into the relations between the types of procedure and the other variables mentioned at the outset of this article, i.e. the reasons for politicization (independent variables)
and the possible contextual factors (moderating variables). In our study, we have shown that the types of procedure have a moderating effect independently from the strategies and interests of the claimants. For at least two more variables, such interaction effects can also be assumed: First, the types of procedure might reflect a certain pattern of legal norms litigated before the court, i.e. depending on the kinds of norms (e.g. rules, standards, or principles) and/or the various contents of those norms (e.g. the different legal fields regulated by constitutional provisions).This would mean that the degrees of politicization vary as a function of the norms or group of norms at hand, which appear systematically different in the different types of procedure. Second, the types of procedure might reflect differing degrees
of publicity, which in turn might be co-determined by the appellants. For instance, types of procedure used by political appellants might have a higher publicity than concrete review procedures.[78]Thus, for the analysis of the impact that types of procedures have on the politicization of constitutional court decisions this study is a first step only.

*     Replication data is available at See instructions at

**   Michael Hein, Post Doc Fellow (Department of Social Sciences, Humboldt-Universität zu Berlin), PhD (University of Greifswald), M.A. (University of Leipzig); Stefan Ewert, Senior Lecturer (Department of Political and Communication Science, University of Greifswald), PhD, M.A. (University of Greifswald). We are grateful to Wiebke Breustedt, Henrike Englert, Maria Haimerl, Chris Hanretty, Sebastian Nickel, Felix Petersen, Susanne Pickel, Hartmut Rensen, and Silvia von Steinsdorff for their helpful comments and suggestions on this article. Additionally, we thank Isabel Mariano Ribeiro, Maja Ulatowski and again Hartmut Rensen for supporting our data research.

[1]        See Alec Stone Sweet, Governing with Judges. Constitutional Politics in Europe (OUP 2000).

[2]        See Dieter Grimm, 'Constitutions, Constitutional Courts and Constitutional Interpretation at the Interface of Law and Politics' in Bogdan Iancu (ed), The Law/Politics Distinction in Contemporary Adjudication (Eleven International 2009), 21.

[3]        See Michael Hein, Stefan Ewert, 'What is "Politicisation" of Constitutional Courts? Towards a Decision-oriented Concept' in Antonia Geisler, Michael Hein, Siri Hummel (eds), Law, Politics, and the Constitution. New Perspectives from Legal and Political Theory (Lang 2014), 31.

[4]       See David P. MacKinnon, 'Integrating Mediators and Moderators in Research Design' (2011) Research on Social Work Practice 675, 678.

[5]        Chris Hanretty, 'The Bulgarian Constitutional Court as an Additional Legislative Chamber' (2014) East European Politics and Societies, 540; Chris Hanretty, 'Dissent in Iberia: The ideal points of justices on the Spanish and Portuguese Constitutional Tribunals' (2012) European Journal of Political Research, 671; Nuno Garoupa, Marian Gili, Fernando Gómez-Pomar, 'Political Influence and Career Judges: An Empirical Analysis of Administrative Review by the Spanish Supreme Court' (2012) Journal of Empirical Legal Studies, 795; Royce Carroll, Lydia Tiede, 'Judicial Behavior on the Chilean Constitutional Tribunal' (2011) Journal of Empirical Legal Studies, 856; Nuno Garoupa, 'The Politicization of the Kelsenian Constitutional Courts: Empirical Evidence' in Kuo-Chang Huang (ed), Empirical Studies of Judicial Systems (Academia Sinica 2009), 149; Christoph Hönnige, 'The Electoral Connection. How the Pivotal Judge Affects Oppositional Success at European Constitutional Courts' (2009) West European Politics, 963.

[6]       Hönnige (fn 5); Stone Sweet (fn 1); Georg Vanberg, 'Abstract Judicial Review, Legislative Bargaining, and Policy Compromise' (1998) Journal of Theoretical Politics, 299.

[7]        Maria Popova, Politicized Justice in Emerging Democracies. A Study of Courts in Russia and Ukraine (CUP 2012); Pilar Domingo, 'Judicialization of Politics or Politicization of the Judiciary: Recent Trends in Latin America' (2004) Democratization, 104.

[8]        See Victor Ferreres Comella, Constitutional Courts and Democratic Values. A European Perspective (Yale UP 2009), 5–8. Using this definition, we examine only the aforementioned three aspects of the constitutional courts' procedural rules. Other procedural elements such as the formation of panels, chambers, or senates, the rules regulating the intra-court negotiations, the execution of oral and/or written proceedings, and the voting rules within the court are left out of consideration.

[9]       Silvia von Steinsdorff, 'Verfassungsgerichte als Demokratie-Versicherung? Ursachen und Grenzen der wachsenden Bedeutung juristischer Politikkontrolle' in Klemens H. Schrenk, Markus Soldner (eds), Die Analyse demokratischer Regierungssysteme. Festschrift für Wolfgang Ismayr zum 65. Geburtstag (VS Verlag für Sozialwissenschaften 2010), 479.

[10]      Steven Schäller, 'Präjudizien als selbstreferenzielle Geltungsressource des Bundesverfassungsgerichts' in Hans Vorländer (ed) Die Deutungsmacht der Verfassungsgerichtsbarkeit (VS Verlag für Sozialwissenschaften 2006), 205.

[11]       See e.g. Nuno Garoupa, Tom Ginsburg, 'Building Reputation in Constitutional Courts: Party and Judicial Politics' (2011) Arizona Journal of International and Comparative Law 539; Wojciech Sadurski, 'Rights-Based Constitutional Review in Central and Eastern Europe' in Tom Campbell, K. D. Ewing, Adam Tomkins (eds), Sceptical Essays on Human Rights (OUP 2001), 315; Helmut Simon, 'Verfassungsgerichtsbarkeit' in Ernst Benda, Werner Maihofer, Hans-Jochen Vogel (eds), Handbuch des Verfassungsrechts der Bundesrepublik Deutschland (2nd edn, de Gruyter 1994), 1637.

[12]      Tom Ginsburg, Zachary Elkins, 'Ancillary Powers of Constitutional Courts' (2009) Texas Law Review 1431; see also Tom Ginsburg, 'Ancillary Powers of Constitutional Courts' in Tom Ginsburg, Robert A. Kagan (eds), Institutions and Public Law: Comparative Approaches (Lang 2004), 225.

[13]      Sascha Kneip, 'Verfassungsgerichtsbarkeit im Vergleich' in Oscar W. Gabriel, Sabine Kropp (eds), Die EU-Staaten im Vergleich. Strukturen, Prozesse, Politikinhalte (VS Verlag für Sozialwissenschaften 2008), 631, 643–647.

[14]      See e.g. Sascha Kneip, 'Constitutional courts as democratic actors and promoters of the rule of law: institutional prerequisites and normative foundations', (2011) Zeitschrift für Vergleichende Politikwissenschaft 131; Hönnige (fn 5). Ali S. Masood, Donald R. Songer, 'Reevaluating the Implications of Decision-Making Models: The Role of Summary Decisions in U.S. Supreme Court Analysis' (2013) Journal of Law and Courts 363, level a similar critique with regard to the state of the art in U.S. Supreme Court studies, which mainly focus on the 'plenary decisions' but leave out the 'summary decisions'. According to that analysis, this narrowing of perspective had created the widespread impression of a conservative court.

[15]      Instead of others, see for the US Supreme Court: Benjamin Woodson, 'Politicization and the Two Modes of Evaluating Judicial Decisions' (2015) Journal of Law and Courts 193; Brandon L. Bartels, Christopher D. Johnston, 'Political Justice. Perceptions of Politicization and Public Preferences Toward the Supreme Court Appointment Process' (2012) Public Opinion Quarterly 105; Stephen M. Engel, American Politicians Confront the Court. Opposition Politics and Changing Responses to Judicial Power (CUP 2011); for Latin America: Diana Kapiszewski, High Courts and Economic Governance in Argentina and Brazil (CUP 2012); Raul A. Sanchez Urribarri, 'Politicization of the Latin American Judiciary via Informal Connections' in David K. Linnan (ed), Legitimacy, Legal Development and Change: Law and Modernization Reconsidered (Ashgate 2012), 307; Carroll and Tiede (fn 5); Domingo (fn 7); for Western Europe: Hönnige (fn 5); Stone Sweet (fn 1); Vanberg (fn 6); Nicos C. Alivizatos, 'Judges as Veto Players' in Herbert Döring (ed), Parliaments and Majority Rule in Western Europe (Campus 1995), 566; for post-socialist Central and Eastern Europe: Popova (fn 7); Michael Hein, 'Constitutional Conflicts between Politics and Law in Transition Societies: A Systems-Theoretical Approach' (2011) Studies of Transition States and Societies 3; for Asia: Jiunn-rong Yeh, Wen-Chen Chang (eds.), Asian Courts in Context (CUP 2014); Björn Dressel, 'Judicialization of Politics or Politicization of the Judiciary? Considerations from Recent Events in Thailand' (2010) The Pacific Review 671; and for Africa: Kierin O'Malley, 'The Constitutional Court' in Murray Faure, Jan-Erik Lane (eds), South Africa: Designing New Political Institutions (Sage 1996), 75.

[16]      Charles H. Sheldon, The Supreme Court. Politicians in Robes (Glencoe Press 1970).

[17]      See Woodson (fn 15); Bartels and Johnston (fn 15); Engel (fn 15); Garoupa (fn 5); Stone Sweet (fn 1); Vanberg (fn 6).

[18]      Manfred G. Schmidt, 'Politisierung' in Manfred G. Schmidt (ed), Wörterbuch zur Politik (3rd ed, Kröner 2010), 630, 630. All German quotes were translated by the authors.

[19]      Instead of others see Pierre J. Schlag, 'Rules and Standards' (1985) UCLA Law Review 379.

[20]     Grimm (fn 2), 27. A recent example of such an only seemingly clear case is a dispute before the Polish Constitutional Tribunal in 2011 concerning the meaning of Article 98, paragraph 2 of the Polish Constitution, according to which parliamentary elections have to take place on 'a non-working day.' To improve their chances in the upcoming elections, the then governing coalition had tried to extend the ballot to two days. This seemingly unequivocal breach of the constitution caused one of the most controversial decisions of the Tribunal, in which no less than eleven (!) of the 15 judges issued dissenting opinions. Judgment of the Polish Constitutional Tribunal, K 9/11, OTK ZU No. 61/6/A/2011.

[21]      Robert A. Dahl, 'Decision-Making in a Democracy: The Supreme Court as a National Policy-Maker' (1957) Journal of Public Law 279, 280.

[22]      For a more comprehensive discussion of this concept of politicization, see Hein, Ewert (fn 3).

[23]      See Bogdan Iancu (ed), The Law/Politics Distinction in Contemporary Public Law Adjudication (Eleven International 2009); Robert C. Post, Neil S. Siegel, 'Theorizing the Law/Politics Distinction: Neutral Principles, Affirmative Action, and the Enduring Legacy of Paul Mishkin' (2007) California Law Review 1473.

[24]     Instead of others see Niklas Luhmann, 'Operational Closure and Structural Coupling: The Differentiation of the Legal System' (1991) Cardozo Law Review 1419; Michael Hein, 'Constitutional Conflicts between Politics and Law in Transition Societies: A Systems-Theoretical Approach' (2011) Studies of Transition States and Societies 3.

[25]      Martina Flick, Organstreitverfahren vor den Landesverfassungsgerichten. Eine politikwissenschaftliche Untersuchung (Lang 2011); Hönnige (fn 5).

[26]     See e.g. Chris Hanretty, 'Judicial Disagreement need not be Political: Dissent on the Estonian Supreme Court' (2015) Europe-Asia Studies 970; Hanretty, Court (fn 5); Hanretty, Dissent (fn 5); Susumu Shikano, Verena Mack, 'Judges' Behaviour and Relationship with Political Parties in a Non-Common-Law Country: the Case of the German Federal Constitutional Court' (2014) Working Paper, Graduate School of Decision Sciences, University of Konstanz 3.8088> accessed 10 December 2015; Santiago Basabe-Serrano, 'Judges without Robes and Judicial Voting in Contexts of Institutional Instability: The Case of Ecuador's Constitutional Court, 1999–2007' (2012) Journal of Latin American Studies 127; Daniel Smilov, The Hybridity of Constitutional Courts – Arbiters in the Absence of Rules (CAS 2009).

[27]      Alexis de Tocqueville, Democracy in America (Regnery 2002 [1835]), 74.

[28]      John Stuart Mill, Considerations on Representative Government (Prometheus 1999 [1861]), 327.

[29]     Garoupa, Ginsburg (fn 11), 550. See also Garoupa (fn 5), 158; Sadurski (fn 11), 318; Simon (fn 11), 1651.

[30]     Tocqueville (fn 27), 75 f.

[31]      Sergiu Gherghina, Sergiu Miscoiu, 'The Failure of Cohabitation: Explaining the 2007 and 2012 Institutional Crises in Romania' (2013) East European Politics and Societies 668.

[32]      Although the impeachment of Richard Nixon (1974) was also politicized, severe violations of the constitution and of the law played a much more important role in that case than in the other two.

[33]      Arnold H. Leibowitz, An Historical-Legal Analysis of the Impeachments of Presidents Andrew Johnson, Richard Nixon, and William Clinton: why the Process Went Wrong (Mellen 2012).

[34]     Upon completing this article, a similar pattern was observable with the ongoing impeachment of Brazil's president Dilma Rousseff, which had begun in late 2015; see 'Brazil's Dilma Rousseff to face impeachment trial' (BBC News, 12 May 2016) accessed 19 May 2016.

[35]      Admittedly, the US Supreme Court is only involved marginally in the procedure, and the parliament not only has the right to initiate but also to decide. Nevertheless, this case offers an illustration of the relevant logic: Even if the final decision would be made by the court (as in many other, particularly European, states), the problem of politicizing a legal matter would remain.

[36]     Uwe Kranenpohl, Hinter dem Schleier des Beratungsgeheimnisses. Der Willensbildungs- und Entscheidungsprozess des Bundesverfassungsgerichts (VS Verlag für Sozialwissenschaften 2010), 47.

[37]      Ex-officiotypes of procedure are not included since there are no applicants whose legal requirements and political opportunities could be described according to the hypotheses. Non-judicial types of procedure such as the observation of elections and referendums and the confirmation of their results, or the registration of political parties (but not: appeals against their non-registration), are not included either, because they are supposed to be not politicized due to their formal character.

[38]      See Allan R. Brewer-Carías (ed), Constitutional Courts as Positive Legislators. A Compa-rative Law Study (CUP 2013); Ginsburg, Elkins (fn 12); Wojciech Sadurski, Rights Before Courts. A Study of Constitutional Courts in Postcommunist States of Central and Eastern Europe(2nd edn, Springer 2008); Ginsburg (fn 12); Georg Brunner, 'Der Zugang des Einzelnen zur Verfassungsgerichtsbarkeit im europäischen Raum' (2002) Jahrbuch des öffentlichen Rechts der Gegenwart 191.

[39]     Another question is, whether these types of procedure are functional and (normatively) desirable in a democratic constitutional state; see Gur Bligh, 'Defending Democracy: A New Understanding of the Party Banning Phenomenon' (2013) Vanderbilt Journal of Transnational Law 1321; Peter Niesen, 'Anti-Extremism, Negative Republicanism, Civic Society: Three Paradigms for Banning Political Parties. Parts I and II' (2002) 3(7) German Law Journal> accessed 19 May 2016.

[40]     Brunner (fn 38), 230.

[41]      Simon (fn 11), 1651.

[42]     See Klaus Stüwe, Die Opposition im Bundestag und das Bundesverfassungsgericht. Das verfassungsgerichtliche Verfahren als Kontrollinstrument der parlamenta-rischen Minderheit (Nomos 1997), 183. Nevertheless, it is again another question whether the abstract review is a functional and (normatively) desirable type of procedure; see ibid, 180.

[43]     See ibid., 170

[44]     Simon (fn 11), 1651.

[45]     See Stüwe (fn 42), 197.

[46]     Katalin Kelemen, 'Dissenting Opinions in Constitutional Courts' (2013) German Law Journal 1345.

[47]     Apart from dissenting opinions, judges are also often allowed to publish concurring opinions. Here, the judge agrees with the decision, but not with the legal reasoning of the judges' majority. Therefore, concurring opinions usually do not indicate political influences.

[48]     Garoupa, Ginsburg (fn 11), 547.

[49]     See Kelemen (fn 46), 1346.

[50]     Hanretty, Dissent (fn 5), 671.

[51]      One example for the first scenario is the decision of the German Federal Constitutional Court concerning the ban of the National Democratic Party in 2003 (BVerfGE 107, 339). In this case, a 3-to-4 minority of the seven judges on the bench decided against the motion (DR=0.57) since for a party ban a two-thirds majority of all eight judges (= 6; one judge was missing) is required. Concerning the second scenario, there are three extreme cases in Bulgaria during the period under investigation (Decisions No. 9/1994, 4/1995 and 10/1997; all decisions of the Bulgarian Constitutional Court are available at accessed 19 May 2016). In these cases, there had been majorities for every part of the decision, but not a single judge agreed on the complete judgment. As a result, the DR reached the maximum value of DR=1.0. In Portugal, we have also observed some judgments with very high DR values in the framework of the second scenario (Judgments No. 423/2008 with DR=0.769, and Judgment No. 338/2010 with DR= 0.833; all decisions of the Portuguese Constitutional Tribunal are available at> accessed 19 May 2016). All mentioned decisions in Bulgaria and Portugal were taken in abstract review procedures.

[52]      Hanretty (fn 26), 970.

[53]      Shikano, Mack (fn 26); Hönnige (fn 5).

[54]     Hanretty, Court (fn 5); Smilov (fn 26).

[55]      Hanretty, Dissent (fn 5); Pedro C. Magalhães, 'The Limits to Judicialization: Legislative Politics and Constitutional Review in the Iberian Democracies' (DPhil thesis, Ohio State University 2003) su1046117531> accessed 19 May 2016.

[56]     Against the background of this knowledge, including these variables in our dataset would not only have been an unreasonably tremendous task of information procurement and coding. In fact, it would have been impossible for more than 99 percent of the decisions of the German Federal Constitutional Court, since they were not taken by one of the two senates (consisting of eight judges each), but by one of the smaller six chambers (three judges). The chambers' decisions are not publicized but only accessible through the official court statistics, which do not provide information on the individual judges' behavior.

[57]      The standard deviation (in brackets) shows that the variability of the DR is rather low in most of the different groups of applicants. The relatively small differences between the applicants in Bulgaria can be explained by the fact that the appointment of the General Prosecutor as well as the judges at the Supreme Court of Cassation and the Supreme Administrative Court (the judicial institutions that are eligible to apply before the constitutional court) is heavily influenced by the parliament. Therefore, the 'highest magistrates' are not as politically independent as their German and Portuguese counterparts.

[58]      Martin Höpner, 'Warum betreibt der Europäische Gerichtshof Rechtsfortbildung? Die Politisierungshypothese'(2010) Sozialer Fortschritt 141, 144.

[59]     German legal scholars therefore speak of 'lines of adjudication' ('Linien der Rechtsprechung'); see Hartmut Rensen, Stefan Brink (eds), Linien der Rechtsprechung des Bundesverfassungsgerichts – erörtert von den wissenschaftlichen Mitarbeitern (De Gruyter 2009).

[60]     For the methodological concept of a moderating variable, see MacKinnon (fn 4). Typical examples for means comparisons in constitutional court research can be found among others in Lawrence Baum, 'Linking Issues to Ideology in the Supreme Court. The Takings Clause' (2013) Journal of Law and Courts 89; Masood, Songer (fn 14) and Schäller (fn 10).

[61]      Excluded from the survey are only decisions in non-judicial and ex officio types of procedure (see n 37), procedural decisions (like on challenges of a judge for the fear of bias), and corrections of previous decisions.

[62]     All countries measured as 'free' according to the Freedom House Index; 'Freedom in the World 2011' (Freedom House) accessed 9 February 2015.

[63]     'Brief History of the Constitutional Court' (Portuguese Constitutional Court) accessed 14 July 2015; see also Joaquim de Sousa Ribeiro, Esperança Mealha, 'Portugal' in Brewer-Carías (fn 38), 721.

[64]     See ibid, 728.

[65]     For the reasons of exclusion of decisions, see n 61. However, the decisions we did not analyze remain relevant with regard to the caseload. When different types of procedure and/or different types of applicants are present in one court decision, this decision and its DR appears several times in our dataset – one data point for each type of procedure and each type of applicant.

[66]     Decisions per year/DR; Germany: Pearsons r=0.109/Significance t=0.649; Bulgaria: r=-0.023/t=0.923; [Portugal: r=0.117/t=0.825]).

[67]     See Donald P. Kommers, Russel A. Miller, The Constitutional Jurisprudence of the Federal Republic of Germany (3rd edn, Duke UP 2012), 65 f.

[68]     See Smilov (fn 26).

[69]     See Magalhães (fn 55), 295.

[70]     The standard deviation (in brackets in table 4) shows that the variability of the DR is rather low in most of the different types of procedure.

[71]      Bundesverfassungsgericht, 'Die Anträge der Bundesregierung, die Verwirkung von Grundrechten auszusprechen, sind nicht hinreichend begründet.' (, 30 July 1996) accessed 19 May. See BVerfG 2 BvA 1/92, 2 BvA 2/92.

[72]      Law on the Federal Constitutional Court (1951) as published on August 11, 1993 accessed 17 June 2015.

[73]      BVerfGE 121, 266.

[74]     BVerfGE 123, 39.

[75]      However, in several cases of other types of procedure we have observed even higher non-aggregated dissenting ratios (see n 51).

[76]     Decisions No. 4/1992 and 1/2000; see Daniel Smilov, 'Constitutionalism of Shallow Foundations: the Case of Bulgaria' in Denis J. Galligan and Mila Versteeg (eds), Social and Political Foundations of Constitutions (CUP 2013), 611, 631–633.

[77]      See Venelin I. Ganev, 'The Rise of Constitutional Adjudication in Bulgaria' in Wojciech Sadurski (ed), Constitutional Justice, East and West. Democratic Legitimacy and Constitutional Courts in Post-Communist Europe in a Comparative Perspective (Springer 2002), 247, 253.

[78]      See for these factors, but with a different leading question, Ulrich Sieberer, 'Strategische Zurückhaltung von Verfassungsgerichten. Gewaltenteilungs-vorstellungen und die Grenzen der Justizialisierung' (2006) Zeitschrift für Politikwissenschaft 1299; Georg Vanberg, The Politics of Constitutional Review in Germany (CUP 2005).


Back to Top