Constitutional Dialogue in the Case of Legislative Omissions: Who Fills the Legislative Gap?

The presence of a legislative lacuna within a legal order has been acknowledged in several European legal orders by explicitly obliging judges to adjudicate legal disputes even in the absence or silence of an applicable legal rule.1 The power of Constitutional Courts to determine that legislative omissions violate constitutional norms and principles can no longer come as a surprise either; the XIVth Congress of the Conference of European Constitutional Courts was even devoted to problems of legislative omissions in constitutional jurisprudence.2 Establishing that an unconstitutional omission is present is one thing, but providing for the necessary redress is another. Only in a few cases will the Court’s annulment or finding of unconstitutionality eliminate this legislative lacuna. After all, petitioners do not want the norm to disappear from the legal order, but rather that its field of application is expanded so that they can benefit from it. Consequently, even more than in the aftermath of other decisions of a Constitutional Court, when legislative omissions are at stake, the response of the other actors, especially legislators and ordinary judges, is crucial and constitutes a remarkable example of constitutional dialogue.3 This is especially the case when the Constitutional Court offers guidelines to these actors or explicitly incites them to fill the legislative gap. Although the case law of the Belgian Constitutional Court (BCC) regarding legislative omissions has already been largely explored, the actual responses by the other actors have not received much attention. For this contribution, I will focus on the legislator’s response to lacuna judgments. After all, it remains the primary responsibility of the legislator to amend unconstitutional legislation, especially in cases where a mere annulment or finding of unconstitutionality does not fill the existing gap. The concept of constitutional dialogue has already been extensively debated. Within a European legal order, the concept is closely related to the interaction between courts, both national and European, and the possible usage of the preliminary reference procedure.4 However, constitutional dialogue is not restricted

Utrecht Law Review | Volume 14 | Issue 1, 2018 to this type of conversation, but describes the nature of interactions between all three state powers in the area of constitutional decision-making, particularly in relation to the interpretation of constitutional rights. 5 In their comprehensive study, published in this journal, Meuwese and Snel provide a (working) definition of constitutional dialogue and offer various approaches for researchers to analyse the concept. 6 According to the authors, a constitutional dialogue is 'a sequel of implicitly or explicitly shaped communications back and forth between two or more actors characterized by the absence of a dominant actor -or at least by a bracketing of dominance -, with the shared intention of improving the practice of interpreting, reviewing, writing or amending constitutions'. 7 They establish two ways in which the concept of constitutional dialogue can be used: either as a lens, or as a method. Regarding the former, the concept is utilized to describe or explain current constitutional arrangements. The empirical or descriptive aspect of this approach relates to the following question: what do we discover about a particular category of public decision-making if we see it as a dialogue? In the normative aspect, research can focus on who we want to be engaging in the dialogue and in what role. 8 If the concept is used as a method, the authors suggest focusing on the organization of the public decision-making processes and the empirical question of what dialogic elements are having what effects. It follows that the normative issue in this approach valuates dialogue as a solution for problems of public decision-making. 9 Meuwese and Snel emphasize that so far little empirical research has been conducted in which the concept of constitutional dialogue is used to describe how different public actors interact. With this contribution I wish to respond, at least partially, to this finding by using the concept of dialogue as a lens and primarily focusing on the descriptive aspect of the approach. Investigating one specific category of public decisionmaking leads us to the following central research question: what do we discover if we see the interaction between the Constitutional Court and the legislator regarding the existence of an unconstitutional gap as a dialogue? To answer this question, the case law of the Constitutional Court and the response by the legislator will be examined in the light of the foregoing definition of constitutional dialogue. 10 I begin by briefly summarizing the most important aspects of the case law of the Belgian Constitutional Court regarding legislative omissions (Section 2) and proceed by explaining the methodological approach (section 3.1) and how the necessary data were gathered, namely the lacuna judgments and legislative responses (section 3.2). Within the next section, the gathered data are tested against the definition provided by Meuwese and Snel. To what extent is an explicit interaction present (Section 4.1)? Which actors are involved in the dialogue (Section 4.2)? Is the Court or the legislator a dominant actor (Section 4.3)? To what extent does the dialogue improve the practice of interpreting, reviewing, drafting or amending the Constitution (Section 4.4)? Finally, I summarize my findings and briefly put forward some preliminary remarks regarding the observed dialogue between the Constitutional Court and the legislator (Section 5).

Case law of the Belgian Constitutional Court
The Belgian Constitutional Court has the power to review federal and subnational acts of Parliament against stipulations which allocate powers between the federal authorities, the Communities and the Regions, and against fundamental rights enshrined in the Constitution. 11 In the case of legislative omissions, the clauses of equality, enshrined in Articles 10 and 11 of the Belgian Constitution, are invoked in the vast majority of cases: an unconstitutional omission will be present when a law excludes one category of persons from its field of application without a reasonable justification. 12 The Constitutional Court, however, does not have the explicit power to review the absence of legislation; the Special Act emphasizes that only a violation by a statute, decree or ordonnance can be reviewed. Consequently, the Court is not competent to review absolute omissions which occur when any legislative provision is absent. 13 Other constitutions and national regulations provide their Constitutional Courts with the possibility to review an absolute omission of legislation. This is for example the case in Portugal, Brazil and Hungary. 14 By contrast, relative omissions which exist when legislation has been enacted but in a partial, incomplete or defective way from a constitutional point of view may be subjected to a review by the Belgian Constitutional Court. It becomes clear that the legislator did not take this competence into consideration when we look at the defined consequences of a finding of unconstitutionality. Article 8 of the Special Act on the Constitutional Court states that if the action for annulment is well founded, the Court shall entirely or partially annul the unconstitutional rule. However, this is not the desired redress; although the equality is restored via a leveldown approach, applicants wish to expand the under-inclusive field of application, rather than annulling the norm in its entirety. The cases where it suffices to declare the exception to a rule unconstitutional in order to expand the field of application are less common. Consequently, the Constitutional Court resorts to some 'creative' dicta to answer questions relating to a legislative lacuna. 15 In particular, the Court uses two different types of decisions to indicate that a legislative omission is present. First, the Court has developed case law in which the contested rule is declared constitutional, but the absence of a similar rule for the excluded category is deemed unconstitutional. The Court often explicitly incites the legislator to introduce a new law, clearly directing the constitutional conversation towards Parliament, usually providing guidance on how this new law should be drafted. 16 Secondly, the Court regularly states that the contested norm violates the Constitution 'to the extent that' certain persons or institutions are excluded from its field of application. Because the Court only declares that the legislative gap is unconstitutional, the contested norm as such remains untouched and can be further applied. Again, this type of ruling does not lead to the desired expansion of the field of application. The Court has resolved this problem by explicitly granting ordinary judges the possibility to fill the legislative gap 'when the finding of unconstitutionality is put in sufficiently precise and complete terms'. 17 Whereas the conclusion that a norm violates the Constitution to the extent that a certain category is excluded still demonstrates the position of the Court as a negative legislator, the explicit empowerment of ordinary judges to fill the legislative gap is clearly an action of the Court as a positive legislator. 18

Methodology
Responding to the above-mentioned findings of Meuwese and Snel, a descriptive research question was formulated: what do we discover if we see the interaction between the Constitutional Court and the legislator regarding the existence of an unconstitutional gap as a dialogue? Consequently, the focus in this contribution is on describing what is going on, primarily in the legislative arena, in the aftermath of a lacuna judgment pronounced by the Belgian Constitutional Court. 20 In order to answer this research question, an empirical approach was adopted. Using the definition of constitutional dialogue provided by Meuwese and Snel, the relevant variables could be identified: How is the decision of the Court formulated? Who are the actors involved in the dialogue? Is the judgment of the Court mentioned within parliamentary proceedings? If so, by whom? How often is the Council of State involved? Who initiates successful legislative responses? How much media attention is given to the judgments of the Court? Is there a legislative reaction in salient cases? If a legislative response is registered, does it adhere to the judgment of the Constitutional Court? Etc. In order to answer the foregoing questions and to collect the necessary data, the case law of the Constitutional Court and the responses by the legislators were analysed. 21

Lacuna judgments of the Belgian Constitutional Court
Providing an exhaustive overview of every lacuna judgment of the Constitutional Court is not feasible. After all, this would imply that more than 3,000 judgments of the Court needed to be examined. What is more, for this research, providing an exhaustive overview is not required. In this study, the focus lies on the dialogue between the Constitutional Court and the legislator. Therefore, I selected those judgments where the Court itself has emphasized the presence of a legislative omission by using keywords such as 'lacuna' and 'omission' to circumscribe the content of the judgment. 22 These keywords are mentioned on the website of the Constitutional Court and in the digital newsletter the Court sends out when new judgments are delivered. In this way, the Court fulfils an important signalling function within the dialogue; the question arises to what extent the legislator will respond when the Court explicitly initiates the constitutional conversation by emphasizing the existence of a legislative omission. This approach has its drawbacks: for example, for 2005 no judgement was marked on the website as containing a legislative omission, although a quick overview of that year's case law indicates otherwise. Considering that I do not strive for an exhaustive overview of all lacuna judgements and focus on the constitutional dialogue initiated by the Court by explicitly stressing the presence of an (un)constitutional omission, the explained approach is justified.
In order to observe enough time between a judgment of the Constitutional Court and a possible change to the unconstitutional legislation, I examined the legislative reaction of judgments from 1985 until 2014. This led to the collection of 108 judgments: 87 were pronounced after a preliminary reference, while 21 were the result of an appeal for annulment. Eventually, within those 108 judgments, the Court pronounced on 116 possible unconstitutionalities, 89 after a preliminary reference and 27 after an appeal for annulment. 23 I refer to possible unconstitutionalities, because the mere fact that the website of the Court indicated a decision as dealing with a legislative lacuna does not necessarily imply that the Court established an unconstitutional legislative omission. What is more, the Constitutional Court uses a wide array of dicta to establish the presence of an unconstitutional legislative lacuna. Sometimes the existence of a legal gap is not even mentioned in the dicta, but in the considerations of the judgment. Consequently, we cannot easily Utrecht Law Review | Volume 14 | Issue 1, 2018 set aside those judgments where the Court dismisses the appeal. Unless explicitly mentioned otherwise, hereinafter I will refer to the 116 (lacuna) rulings of the Court.

Legislative response
Neither the federal Parliament, nor the Parliaments of the Communities or Regions provide for a systematic follow-up or a specific parliamentary proceeding to comply with the case law of the Constitutional Court. 24 This complicated the task of retrieving the legislative response in a structured manner. Consequently, I resorted to an ad hoc solution. First, I assessed whether the contested norm had been amended. Secondly, for decisions taken from 2011 until 2013, the report of the Parliamentary Committee for the Evaluation of Legislation provided useful insights. 25 Figure 1 shows the legislative response, or the lack thereof, to the 116 rulings of the Court. For 76 rulings, a legislative response was detected. Taking a closer look at those rulings to which the legislator has not yet reacted, we find 9 rulings that do not necessarily require a legislative response; these are rulings where the Court declares that it is incompetent to review the case, where the Court dismisses the appeal without establishing anywhere in the judgement that a legislative lacuna is present, where the question does not need to be answered because the Court is merely asked to offer guidelines to ordinary judges and rulings where the Court fills the legislative gap by a mere annulment. Finally, the figure shows that for 31 rulings no legislative response is registered. Only in 2 out of these 31 rulings did I find explicit considerations by the legislator that it would not take any steps to amend the legislation.

Dialogue as a lens
The definition provided by Meuwese and Snel contains four distinct aspects: implicit or explicit communications, two or more actors, the absence of a dominant actor and the intention of improving the practice of interpreting, reviewing, writing or amending constitutions. 26

Explicit dialogue
This research focusses on the explicit dialogue which unfolds after the Constitutional Court has determined whether or not a legislative omission is present. The reason for this demarcation is primarily a methodological one. I selected those cases of the Constitutional Court where it explicitly emphasized the presence of an (un)constitutional omission by using keywords such as 'lacuna' and 'omission' to describe the content of the case. Furthermore, in order to analyse the possible reaction by the legislator, I examined whether the contested norm had been amended and whether the judgment of the BCC was mentioned in other parliamentary proceedings or means of political control. Purely technically, it is very difficult to search for the possible implicit dialogue that might have taken place. This would mean that for every lacuna 24 The federal Parliamentary Committee for the Evaluation of Legislation performs in a substandard manner and, consequently, this procedure cannot guarantee a thorough examination of the legislative response. 25 Cf. Section 4.2.4, infra. 26 Meuwese & Snel, supra note 6, p. 126.
Utrecht Law Review | Volume 14 | Issue 1, 2018 judgment, parliamentary proceedings and documents regarding the mere subject of the judgment need to be examined. Considering that some preliminary judgments were decided more than two decades ago, such an inquiry is not feasible.
Hereinafter I will discuss to what extent we see such an explicit dialogue taking place between the Constitutional Court and the legislator. It is important to note that from the mere fact that the legislator did not amend the contested norm or did not explicitly refer to the decision of the Court, we cannot deduce that an (implicit) dialogue is absent. On the other hand, when the decision of the Court is followed by some legislative action, this does not necessarily mean that the judgment of the BCC was taken into consideration. The mere fact that the contested norm was amended does not necessarily imply the existence of a reaction to the decision of the Constitutional Court. 27 4.1.1 Case law of the Belgian Constitutional Court By bringing an action before the Constitutional Court, either via an appeal for annulment or via a preliminary question, the constitutional dialogue is clearly initiated. As stated before, the Constitutional Court is only competent to review relative omissions, i.e. existing legislation that proves to be insufficient or incomplete. The Constitutional Court resorts to different dicta to determine whether an unconstitutional legislative lacuna is present or not. The question arises whether the way in which the Court determines that a legislative omission is present may influence the legislator's response. The Court clearly engages in the dialogue when it resorts to the explicit consideration in its judgment that the absence of a legislative norm for the excluded category is unconstitutional, especially when it explicitly incites the legislator to react; the Court resorted to this consideration in 31 rulings. However, the research shows that the use of this formulation does not necessarily provoke a legislative response; the legislator reacted in 17 out of the 31 rulings (55%). Moreover, in those 31 rulings the Court emphasized that the unconstitutionality was not to be found in the contested norm, but rather in the absence of a similar norm for the excluded category. Remarkably, for 12 out of the 17 rulings where the legislator reacted, it did amend the contested norm although, according to the Court, that norm was nevertheless constitutional (71%). For 4 rulings, another existing legal norm was amended and in only 1 case did the legislator introduce a new legal norm. The Court and the legislator seem to have a different view on where the legislative omission is located.
In 35 rulings, the Court explicitly granted ordinary judges the competence to fill the unconstitutional legal gap. Although this is also an example of an explicit dialogue, it is not directed towards the legislator, but towards ordinary judges. We could assume that in these cases the legislator would not find it necessary to change the contested legislation, because ordinary judges may provide the necessary redress. This argument, however, is not supported when we look at the correlation between the legislative responses and the usage of this type of consideration: 25 out of the 35 rulings were followed by a legislative response (71%). Apparently, the legislator prefers to fill the unconstitutional gap via a legislative process instead of a judicial solution.
Consequently, a paradoxical result stems from the foregoing: the legislator's response rate is higher when the Court makes it possible for ordinary judges to fill the legislative gap than when the Court explicitly incites the legislator to respond. A possible explanation for this result could be the fact that rulings where the Court protects the margin of appreciation of the legislator by emphasizing that only it can fill the unconstitutional gap are often more (politically) sensitive cases and thus where it is more difficult to find a (political) legislative solution. Cases in which the Court clarifies how ordinary judges may fill the legislative gap are often less (politically) sensitive, allowing the legislator to transpose the solution offered by the Court into a piece of legislation without much debate. 28 Utrecht Law Review | Volume 14 | Issue 1, 2018

Legislative response
Beforehand it is important to note that there is no time limit within which courts must submit preliminary questions to the Belgian Constitutional Court. Consequently, the norm applicable to a specific case and, as such, the subject of a preliminary question by the ordinary judge may have already been changed. For example, in 2012 the Court of Appeal of Brussels needed to settle a dispute regarding civil liability in the case of an accident that happened in 1990. 29 The Court of Appeal asked the Constitutional Court to decide on the constitutionality of a legislative norm which was applicable in the case at hand, but which had already been amended more than 20 years earlier. In total, out of the 76 legislative actions, 15 entered into force prior to the decision of the Constitutional Court; 30 the legislative action was by no means a 'reaction' to the judgement of the Court.
On the contrary, looking at the interaction between the Constitutional Court and the legislator as a dialogue, we can assume that when the legislator explicitly refers to the decision of the Court during parliamentary proceedings, this constitutes an actual response to the dialogue initiated by the Court. 31

Explanatory memorandum 32
Only an explanatory memorandum 4 With other parliamentary proceedings 14 With the advice of the Council of State 3 Combination of other parliamentary proceedings and the advice of the Council of State 11 Only other parliamentary proceedings 6 Only in the advice of the Council of State 2 Other parliamentary proceedings and the advice of the Council of State 3 Total 43/50 Consequently, I examined the parliamentary proceedings for 61 legislative reactions. 32 I further differentiated between legislative actions that were initiated prior to or after the Court's ruling. In 11 cases, the bill to change the legislation had already been introduced in Parliament prior to the Court's judgment, but had not yet been adopted. Considering that the explanatory memorandum is attached to the bill when it is introduced in Parliament and the advice of the Council of State is given even prior to the introduction of a bill in Parliament, it is only logical that in these documents no reference is made to 29 BCC n° 31/2013. 30 One of these laws had not yet entered into force, but had already been passed in Parliament, BCC n° 35/2007. Two legislative responses in the aftermath of case BCC n° 185/2002 are not included in this number of 15; the relevant bill was introduced in Parliament after the judgment of the Court, but it entered into force prior to the Court's ruling due to the retroactive effect that was given to it. Consequently, the response can still be seen as an actual 'reaction' to the Court's ruling. 31 The reference to parliamentary proceedings also includes the letter to the King that the executive uses to justify its performance. 32 Two of these reactions had already been introduced in Parliament, but had not yet been adopted (BCC n° 129/2010 and BCC n° 133/2013).
Utrecht Law Review | Volume 14 | Issue 1, 2018 the judgment of the Court. Nevertheless, we can take these legislative responses into account because the opportunity remained for the legislator to introduce the decision of the Court into the parliamentary debates, i.e. in further amendments, reports, etc. However, the research shows that when a bill had already been introduced in Parliament, only in 2 out of 11 cases (18%) was the judgement of the Court discussed in proceedings other than the explanatory memorandum. By contrast, when the legislative response was initiated after the pronouncement of the Court's judgment, Figure 2 shows that in no less than 43 out of 50 legislative responses (86%) the ruling of the Constitutional Court was explicitly mentioned during parliamentary debates. From the mere reference to a judgment of the Court, we cannot assume that a thorough assessment of the Court's decision took place. When we take a closer look at the considerations of the legislator, we see that it often merely cites sections of the Court's judgments or gives a brief summary of the decision of the Court and states that the respective bill is intended to fill the determined unconstitutionality. The fact that in only a few cases could a substantial analysis of the merits and demerits of a judgment be seen in order to justify the legislative response 33 does not necessarily imply that this assessment did not take place in other cases.
The constitutional dialogue cannot be limited to successful legislative responses. The unsuccessful initiatives equally demonstrate the conversation between the Court and the legislator, albeit with a different result. In reaction to 23 rulings of the Court, 54 legislative drafts were unsuccessfully initiated in Parliament, each of these explicitly dealing with the rulings of the Court. A notable example concerns the legislative undertakings that followed after the instruction by the Court that the legislator should define the conditions and procedures which are necessary to grant parental authority to persons who do not have an established affiliation with the child, e.g. new partners of the parents. 34 After three bills were introduced in Parliament, the fourth was successful and changed the rules regarding the adoption of children. This amendment, however, was deemed insufficient and 8 new bills were introduced in Parliament to establish a new legal concept of parenthood.

Legislator in a broad sense
As mentioned in the introduction, the constitutional dialogue takes place between all three state powers, but for this contribution I focus on the conversation between the Constitutional Court and the legislator defined in a broad sense; not only bills from Parliament, but also decrees from the executive and even circulars from the judiciary or Federal Departments are taken into account. For example, the circular of the Board of Attorneys General of the Courts of Appeal provided for the necessary redress after decision n° 60/2012. Prior to the judgment of the Court, the legislator had already adopted a bill to change the contested criminal legislation, but it kept postponing the norm's entry into force. 35 Within our data, two unconstitutional omissions were amended by a decree from the Government. Prior to the Court's ruling, the Government of the Brussels-Capital Region had provided for the specific criteria necessary to verify whether a project needs to be subject to an environmental impact assessment. 36 By Royal Decree the federal Government filled the unconstitutional absence of a norm that imposed the same tax exemption on the National Railway Company of Belgium (NMBS) as was already imposed on the railway infrastructure manager.

Different actors within Parliament
Considering that the Belgian Constitutional Court is competent to review federal and subnational acts of Parliament, it cannot come as a surprise that the majority of the legislative reactions after a judgement of the Court follows from a bill passed in Parliament. Within the Belgian legal order, the legislative function is assigned to the legislature, consisting of the King, the House of Representatives and the Senate. It needs to be emphasized that in the legislative process the role of the King is merely formal and, in fact, it is the Government that acts. 38 On the Belgian federal level, Ministers of the federal Government and Members of Parliament, i.e. Members of the House of Representatives, as well as Senators, have the power to introduce bills in Parliament. 39 Figure 4 shows that out of 74 successful legislative actions, 40 53 were Government bills (71%). Consequently, the Government plays a pivotal role in actually filling the unconstitutional legislative gap. If we compare this with the unsuccessful initiatives mentioned above, we see that no less than 53 thereof were proposals introduced by Members of Parliament, whereas only one Government bill was not adopted in Parliament. These numbers may not be surprising and they reflect the normal course of business within the Belgian legislative procedure; 90% of the laws that are adopted in Parliament follow from Government initiatives, although the number of proposals initiated by MPs exceeds the number of Government bills. 41

Successful bill
Unsuccessful bill

Figure 4 Bills introduced in Parliament
The foregoing does not imply that the role of Members of Parliament (MPs) is inferior within the constitutional dialogue. First, whereas the successful amendments are mostly part of a larger reform or a so-called programme act, i.e. a statute that contains a great variety of subjects, the unsuccessful proposals by the MPs are often introduced with the sole intention of filling the unconstitutional legislative gap as determined by the Court. Secondly, MPs can direct their attention towards legislation by means of political control; especially the possibility to ask written or oral parliamentary questions to the competent Minister or State Secretary will strengthen the position of Parliament as an important actor within the constitutional dialogue. 42 Sixteen lacuna judgments of the Constitutional Court have been the subject of parliamentary questions. 43 The questions show that MPs often have a good understanding of the judgment and its possible consequences; 44 MPs explicitly ask the Minister or State Secretary what actions he or she will take in order to comply with the case law of the Court. Although this type of political control can be an important incentive for the Government to amend the unconstitutional legislation, the research shows that for 8 out of the 16 judgments, still no legislative response is registered. 38 Popelier & Lemmens, supra note 11, pp. 120-121. 39 We will not discuss the respective competences of the House of Representatives and the Senate which have changed significantly since the constitutional reform in 2014. It suffices to say that the competence of the Senate to introduce bills has been significantly diminished, but this does not impact this research seeing that it examines legislative responses prior to the constitutional reform in 2014. 40 The two legislative reactions by decree are not included because they are not introduced in Parliament. 41 47 According to the legislator, it sufficed to amend only one of these provisions in order to eliminate the two instances of discrimination. The Council of State however emphasized that for purposes of legal certainty, a further clarification was required in the other provision. 48 The legislator did not follow the advice of the Council and adopted the statute without further amendment. Figure 2 shows  Although the legislation regarding the establishment of this committee was adopted in 2007, it did not actually sit until 2011. According to Article 9 of the Act establishing the Committee, it takes into consideration, once a month, the decisions of the Constitutional Court that may influence the effective functioning of the legal system. 51 From these considerations, a report will be drafted and, if consensus is reached, a bill will be introduced in Parliament in order to eliminate the legislative problems that emerged. 52 Consequently, this Committee holds a crucial position within the dialogue when it comes to legislative lacunae; it has the primary task of monitoring unconstitutional omissions as indicated by the Constitutional Court and, by initiating legislative amendments, it can provide the necessary redress. Looking at the output of this Committee, however, it becomes clear that it does not live up to this expectation, on the contrary. In a timespan of five years, the Committee has only delivered two reports in which the case law of the Constitutional Court was examined. 53 In comparison, the Attorney General at the Belgian Supreme Court (Court of Cassation) and the Board for Attorneys General at the other courts have delivered yearly reports to the Committee since 2007, i.e. years before the Committee first sat, providing an overview of the legislation that contains difficulties in being applied or interpreted by ordinary courts. 54 The evaluations by the Committee only led to one proposal that was introduced and adopted in Parliament in order to 45  ameliorate several different statutes. 55 One of the members of the Committee highlighted that the proposal introduced namely technical amendments that did not require a political stance to be taken on the subject. 56 Considering that in order to fill an unconstitutional legislative omission, a mere technical omission will often not suffice, the Committee will not be able to ensure the necessary redress via the legislative process.
The political control carried out by the Committee is insufficient to carry any weight in the legislative reaction. The Committee does not examine the decisions of the Constitutional Court itself. It addresses the competent parliamentary commissions, asking if the case law of the Court has been examined, what the conclusion of the examination was and if a bill will be introduced in order to remedy the difficulties highlighted by the Court. The answers of the commissions and Ministers are presented in charts and tables, clustering the cases of the Court into the following categories: no legislative response needed, bill in preparation, bill submitted to Parliament, cases under examination and cases that resulted in a legislative amendment. Therefore, these reports form an overview of the current state of affairs as uttered by the parliamentary commissions or Ministers, rather than an actual and thorough assessment of the (absence of the) legislator's response. Only once has the committee uttered its own opinion that the unconstitutional norm should be amended to fill the legislative gap. 57 Moreover, the answers of the Ministers are not necessarily sufficient, as is demonstrated in the aftermath of a ruling pronounced by the Constitutional Court in 2010. 58 Contrary to sanctions imposed by a criminal judge, when a labour judge confirms a sanction by the social services to suspend the payment of a living wage, he/she cannot postpone the enforcement of the sanction. The Constitutional Court found that the absence of a legislative norm authorizing the labour judge to postpone the enforcement of the sanction violates the constitutional principles of equality and non-discrimination, as enshrined in Articles 10 and 11 of the Belgian Constitution. It explicitly emphasized that a legislative reaction was needed to indicate under which conditions the labour judge could grant such a postponement. The legislator, however, did not react. Consequently, and indeed remarkably, given the absence of a legislative amendment, the judge who made the preliminary reference to the Constitutional Court in the first case decided to refer it a second time, hoping that this time the Court would allow the referring judge and others to fill the legislative gap. 59 The Court, however, repeated its previous ruling, emphasizing that the legislator should take the appropriate steps. According to the report of the Parliamentary Committee, the Minister did not follow the Court's reasoning and believed that no legislative response was required because the social services have the power not to impose the sanction of suspending payment in certain circumstances. 60 This answer, however, does not suffice; the competences of the social services are unrelated to the power of the labour judge. The second preliminary reference by the labour judge clearly demonstrated this inadequacy.

Media coverage as an influential factor
Hogg and Bushell have determined that judicial decisions can cause a public debate in which Charter (or constitutional) values play a more prominent role than they would if there had been no judicial decision. 61 One way to determine whether such a public debate has taken place is by looking at the media coverage of judicial decisions. Recent studies, mostly on the US Supreme Court, focus on the characteristics of 55  cases which trigger media attention and consequently influence public opinion. 62 For this contribution, I apply a different approach, thereby using the concept of constitutional dialogue as a lens: what do we discover when we look at the interaction between media coverage and the legislative performance? 63 Put differently: can it be discerned that the public debate is (explicitly) taken into account by the legislator? For this analysis, I used the dataset of De Jaegere regarding media attention for decisions of the BCC. She relied on data collected by the library and documentation office of the Belgian Constitutional Court (last updated January 2017). These data reflect how many articles on each individual judgment appeared in the newspapers, i.e. newspapers that have the widest circulation in Belgium, either in the Flemish or Walloon region (e.g. De Standaard, De Morgen (Flemish Region); Le Soir, La Libre Belgique (Walloon Region)). The data were collected on the basis of specific keywords, such as 'constitutional court' and 'judicial review'. 64 Overall, the dataset demonstrates that very little news media attention is given to judgments of the Belgian Constitutional Court. Looking at all judgments in which the Court gave a final decision on the substance of the case up until 2015 (n=3145), 73% of the BCC's judgments were not mentioned in any newspaper article, 65 22% of all judgments were mentioned in 1 to 5 newspaper articles 66 and only 5% of all judgments of the Belgian Constitutional Court were mentioned in 6 or more newspaper articles. 67 Concentrating on the 108 lacuna judgments, we see the following results regarding media coverage. 68 For 3 cases, the newspapers only referred to the pending appeals or references to the Constitutional Court, 5 judgments received attention not only prior to, but also after the decision and 27 judgments were discussed after the pronunciation of the decision by the Court. Consequently, 73 lacuna judgments were not the subject of any public debate (68% of all lacuna judgments). This number is not surprising when we compare this with the above-mentioned conclusion that, in general, 73% of all judgments of the BCC are not mentioned in any newspaper. The highest amount of articles was 14 for lacuna judgment n° 11/2009 (which constituted a mere annulment of the contested norm), followed by 6 articles for lacuna judgment n° 154/2007 and 5 articles for lacuna judgment n° 81/2008. All other lacuna judgments were covered in 4 newspaper articles or less (see Figure 5).

Figure 5 Lacuna judgments reported in the media
What is more, we might assume that when the Court explicitly declares that the absence of a legislative norm is unconstitutional, the Court sends a clear message to the legislator and to other actors that a legislative reaction is required. Thus, the media coverage of these cases is expected to be higher. However, the explicit consideration that the absence of a legislative norm is unconstitutional does not guarantee more media coverage and, consequently, a larger public debate; of the 31 judgments where the Court pronounced a ruling in this way, only 9 were subsequently discussed (29%) in a total of 19 newspaper articles. Consequently, 71% of these judgments did not lead to a public debate. These numbers are consistent with the research by De Jaegere that shows that, in general, so-called modulated outcomes are considered less newsworthy than simple invalidations. 69 The question arises to what extent the legislator explicitly takes into account the ongoing public debate when bills are drafted. To answer this question, I looked at those 32 lacuna judgments that were discussed in the media and I examined if these newspaper articles were explicitly mentioned in the explanatory memoranda. After deducing those judgments where the legislative amendment had already entered into force or where the explanatory memorandum had already been initiated prior to the judgment of the Court, those judgments where no legislative response was necessary or simply no legislative reaction whatsoever was registered, the explanatory memoranda for 20 judgments were analysed. I used the keywords 'press', 'media', 'newspaper' and 'news', but in none of the examined explanatory memoranda was a reference to these keywords made. 70 Consequently, an explicit reference to the public debate as justification for a legislative reaction was not found. This does not mean that such a consideration was not taken into account by the legislator, but it demonstrates that for these lacuna judgments the legislator did not explicitly involve society as an actor within the constitutional dialogue.

Absence of a dominant actor
When one of the actors holds a dominant position in the conversation, not a dialogue but a monologue would arise where the other actors simply follow the constitutional interpretation imposed by the dominant actor. 71 In this regard, Hiebert defines a new parliamentary rights model, rejecting the dichotomy between judicial supremacy and parliamentary sovereignty when it concerns the responsibility for assessments about rights. 72 Hiebert's model is based on two principles. First, fundamental rights are not merely protected via an ex-post evaluation by the courts, but also Ministers and Members of Parliament have the duty to review rights distinct from, and prior to, judicial review. Secondly, the author emphasizes that via this model, an 69 De Jaegere, supra note 64, pp. 151-152. 70 In Dutch these keywords are respectively 'pers', 'media', 'krant' and 'nieuws'. 71  intra-institutional and interinstitutional reflection on rights is warranted; the expertise of both actors will nourish the dialogue and lead to high quality legislation and a better protection of fundamental rights. 73 Hereinafter I will examine in a descriptive manner to what extent the Belgian Constitutional Court respects the margin of appreciation of the legislator (4.3.1), when the BCC has changed its case law to provide the necessary redress (4.3.2), to what extent the legislator follows the instructions given by the BCC (4.3.3) or (deliberately) deviates therefrom (4.3.4). To conclude, three salient cases are examined to distinguish the role of both actors in politically-sensitive cases (4.3.5).

Deference to the legislator
The Belgian Constitutional Court demonstrates great deference to the legislator's prerogatives. First and foremost this becomes apparent in those 31 rulings where the BCC decided not to invalidate the contested norm, but determined that the absence of a legislative norm was unconstitutional. By emphasizing that it is for the legislator to fill the unconstitutional gap, the BCC upheld the legislative prerogatives. Important to note in this regard are the two rulings of the Court where it concluded that the absence of a legislative norm authorizing the labour judge to postpone the enforcement of a sanction violated the constitutional principles of equality and non-discrimination. In these two judgments the Court ruled that it was only up to the legislator to decide under which conditions a postponement can be granted and to determine under which conditions and according to which procedure such a postponement can be withdrawn. 74 Even when the labour judge and public prosecutor emphasized the continued absence of a legislative reaction, and therefore referred the second preliminary question to the Constitutional Court, the latter adhered to the legislator's prerogatives by confirming the need for a legislative reaction. After all, the Court could have provided for the necessary redress by formulating its ruling as a constitutional violation 'to the extent that' no similar possibility for a postponement was granted to the labour judge and allowing ordinary judges to fill the legislative gap.
Another way in which the BCC shows deference towards the legislator is by modulating the retroactive effect of its judgments. This temporal effect can entail grave financial and administrative consequences, thereby infringing the principle of legal certainty. 75 The Court decided to mitigate these harsh consequences in 4 out of the 116 rulings, in 3 of these cases it provided the necessary time for the legislator to amend the unconstitutional legal gap. In this way the BCC uses a dialogic device to allow the legislator to select among the range of options that would satisfy the Constitution. 76

Wilful Constitutional Court
However, the BCC does not always protect the legislator's prerogatives by adhering to its previous case law in which it states that it is up to the legislator to provide the necessary redress. This becomes clear in two cases regarding the position of recognized stateless persons where the Court reversed its previous ruling. Merely being recognised as a refugee automatically confers a right of residence in Belgium. Recognition as a stateless person, however, does not automatically entail such a consequence. In its ruling n° 198/2009 the Court stated that the unconstitutionality did not follow from a shortcoming in the rule regarding refugees, but that it was enshrined in the absence of a similar norm for recognized stateless persons. Approximately two years later, the Court explicitly emphasized the absence of a legislative response and consequently changed the formulation of its ruling; the Court determined that the existing rule violated the principles of equality and non-discrimination to the extent that it does not provide recognized stateless persons with a similar right of residence as provided for refugees. The Court even granted ordinary judges the power to 73  fill the legislative gap, more precisely by assuming a right of residence and allowing the benefit of family support to people recognized as stateless persons. 77 By granting ordinary judges the possibility to fill the legislative gap, the BCC takes matters into its own hands by providing the necessary redress in concrete pending and future cases. In this way the Court obliges the legislator to react when it does not agree with the solution offered by the Court; after all, until a legislative amendment is made, ordinary judges will follow the instructions given by the Constitutional Court.
What is more, in 2014 the BCC further departed from the division between judgments where the absence of a legislative norm was found unconstitutional, thus implying that the legislator needs to react, and judgments where it declares the contested norm unconstitutional 'to the extent that' a certain category is excluded, thereby instructing the ordinary courts to fill the gap. In that specific case, the BCC had to rule upon the difference in the calculation of holiday pay. 78 According to one regime, sick days were taken into account to calculate the amount of holiday pay, but in the other regime, this was not the case. Municipalities could choose between the two regimes. The Constitutional Court found the absence of a corrective norm which would eliminate the difference in calculation to be unconstitutional. This time, the Court did not refer to the legislator's prerogatives to fill this unconstitutional gap, but it explicitly instructed the ordinary courts to fill the legislative gap and to end the violation of the principles of equality and non-discrimination. 79 Seeing that the unconstitutionality is not located in the specific contested norm, but in a much broader formulated 'absence' of legislation, the BCC offers ordinary courts a great deal of leeway to fill the legislative gap.

Adherence to the opinion of the Belgian Constitutional Court
The examination of the case law of the Belgian Constitutional Court during parliamentary debates proves that the legislator often explicitly takes into account the rulings of the Court. 80 Moreover, despite the fact that the legislator is free to amend the unconstitutional norm as it chooses, 81 it generally changes its legislation according to the case law of the Court. In 72 out of the 76 identified legislative responses, the legislator adhered to the principled opinion uttered by the BCC. Consequently, the Constitutional Court plays a crucial role in the delineation of constitutional principles. The risk, however, is that this would encourage the legislator to defer to judicial statements about rights rather than to engage in an independent consideration of the meaning of constitutional values. 82 Moonen stated that if the considerations of the Belgian Constitutional Court would be accepted without any criticism, this would lead to an impoverishment of the debate regarding our fundamental values. 83 Hereinafter we will see that the legislator does not always follow the directions given by the Constitutional Court.
It must be added that in order to see how the constitutional dialogue further develops, it must be examined whether the legislative reactions to the lacuna judgments were subject to an appeal for annulment or a preliminary reference. 84 In this way it can be discerned how the Court deals with legislation which is intended to implement a Court's ruling.

Wilful legislator
The following three findings demonstrate, in turn, that the legislator does not blindly follow the case law of the BCC. First, in 4 out of the 76 legislative responses, the legislator did not fill the legislative gap as instructed by the BCC. In two rulings, the Court opted for a level-down approach, although the Court declared the contested beneficial norm to be unconstitutional to the extent that a certain category could not profit therefrom. 85 For the remaining two legislative responses, we might question whether the legislative amendment has actually eliminated the unconstitutionality. 86 We refer to the case where the Court decided that the contested norm was unconstitutional to the extent that the parent who was not the beneficiary of child support was unable to submit a claim before the labour court in order to receive additional child support. 87 The contested norm was amended, specifying who was competent to receive the additional child support and under which circumstances. However, the possibility for the parent to bring his claim before the labour court was not explicitly determined. In its advice, the Council of State emphasized that the legislator did not meet the requirements as set forth in the judgment of the Constitutional Court. 88 The legislator responded that against the decisions taken by the Child Support Fund, an appeal can be lodged with the labour court, but it refused to explicitly amend the legislation in this way. By refusing to explicitly subscribe to this right of action, the legislator prioritised its own view and set aside the interpretation by the Court.
Secondly, it became clear that even when the Court determines that the contested norm is constitutional, but the absence of a similar norm is not, the legislator mostly changes the contested norm. 89 In this way the legislator adheres to the principled opinion of the BCC by expanding the field of application, but it does so in a different (formal) manner than the BCC suggested.
Finally, it has to be emphasized that for 31 rulings no legislative response has, as yet, entered into force. Although for 7 of these rulings, proposals and Government bills were (unsuccessfully) introduced in Parliament, 90 the legislator ends the dialogue with the Court when no bill is adopted. Moreover, for 2 rulings the legislator explicitly emphasized that it will not react. 91 These two cases were considered before when discussing the function of political control (or the lack thereof) which the Parliamentary Committee for the Evaluation of Legislation can perform. The Minister refused to attribute to the labour judge the competence to postpone the suspension of the payment of a living wage. Given that the social services have the power not to impose the sanction of suspending payment, no legislative reaction was deemed necessary. 92 As stated before, this answer does not suffice, as clearly demonstrated by the second preliminary reference by the labour judge.
It is important to note that in 10 out of these 31 rulings where no legislative response has yet been registered, the Belgian Constitutional Court granted ordinary judges the explicit competence to fill the legislative gap. In this way the BCC partially mitigates the consequences of the absence of a legislative response. However, this also means that for 21 out of 107 lacuna rulings (20%) where a response is required, an unconstitutional omission is still present within the legal order, without any form of redress being offered.

Response in salient cases
Some might suggest that the absence of a legislative reaction after a certain number of rulings of the BCC does not paint the whole picture of the constitutional dialogue. More specifically it will give useful insights if the dialogue between the BCC and the legislator is uncovered in cases where the stakes are politically high. Using the characteristics of participation by a large, diverse group of litigants, media coverage during the decision-making procedure and a deliberation in plenary session, De Jaegere deduces 57 highly salient cases 85  within the case law of the Belgian Constitutional Court. 93 If we compare this list with the list of 108 lacuna judgments, three highly salient lacuna judgments arise. All three cases are appeals for annulment and each judgment has more than one ruling on the presence of unconstitutional legislative omissions. Content-wise, but also regarding the types of adjudication used by the BCC, there is a big difference between these three judgments. However, this does not detract from the fact that all unconstitutional legal gaps were eventually filled, either by the Constitutional Court or by the legislator in accordance with the guidelines provided by the Court. What follows is a quick overview of the interaction between the BCC and the legislator in these three cases. 94 In the first case, an appeal for annulment was brought before the BCC against the Weapons Act which had been restricted after a racial incident involving gun violence in Antwerp. Arms dealers, arms collectors and people who exercise a profession that necessitates the use of a firearm need to have official authorisation. In the first ruling, the BCC declared a provision unconstitutional to the extent that the continued possession of firearms was not included as a legal reason to apply for authorisation. The legislator amended the contested norm within 8 months, adding the legal reason and even explicitly mentioning the judgment of the Court in the amending legislation: 'in Article 11 of the same law, partially annulled by judgment n° 154/2007 of the Constitutional Court, the following changes are made: (…)'. 95 Secondly, the Weapons Act provided officials with unlimited access, at any time, to all locations where these authorised persons conduct their activities. The Court found the absence of any safeguards for the rights of persons under investigation, such as intervention by an independent judge or a limitation on house searches with regard to locations or hours of the day, disproportionate to the pursued aim of public safety. It therefore annulled the far-reaching competence of the officials, making the general rules regarding the inviolability of homes applicable. To mitigate the grave consequences for ongoing investigations, the BCC mitigated the retroactive effect of its annulment. The annulment of the far-reaching competence immediately eliminated the unconstitutional absence of safeguards and the Minister explicitly decided that no further legislative reaction was needed. 96 The second case related to the provision of energy. 97 First, the BCC annulled the term 'natural' in the Electricity Act that limited the prohibition on the accumulation of participation in certain undertakings to natural persons. In this way, Directive 2009/72/EC was not being correctly implemented, but by annulling the term 'natural', an immediate expansion of this prohibition to legal persons took place. Secondly, the Court found a norm in the Electricity Act to be unconstitutional to the extent that it did not oblige the publication of certain index formulas for SMEs. The BCC granted ordinary judges the competence to fill the legislative gap and, within 5 months, the legislator had expanded the obligation to publish to SMEs. Thirdly, the Court found an unconstitutional omission in the Electricity Act to the extent that a provision was not applicable to operators who are not linked to the transmission networks. Again, the BCC used the consideration that ordinary courts could fill this legislative gap. Meanwhile, the legislator had already filled the unconstitutional gap 7 months prior to the decision of the Constitutional Court. The latter even referred to this legislative amendment in its judgment.
The last case relates to the topic of family reunification for migrants in which the BCC established four unconstitutional legislative gaps. 98 The Court declared one of the contested norms to be unconstitutional to the extent that the possibility to lower the age limit for family reunification with a person from a third state was not applicable for family reunification with an EU citizen. The BCC did not declare the ordinary courts competent to fill the legislative gap, nor did it explicitly emphasize that a legislative reaction was required. In a second ruling, the BCC declared a norm unconstitutional to the extent that no procedure for family reunification was provided for family members who are not partners or blood relatives, but who 93 De Jaegere, supra note 64, pp. 100-128. 94 For this contribution I will only discuss those aspects of the judgments that involve legislative omissions. 95  are dependent on the EU citizen in question. This time the BCC explicitly mentioned that it was up to the legislator to fill the legislative gap. The same considerations were made in the third ruling where the BCC annulled the contested norm to the extent that it did not provide for an exception to the requirements of resources when the applicant requests reunification with his underage children or the underage children of his partner when the partnership is deemed equal to a marriage in Belgium. Finally, the BCC found the absence of a legislative norm which would allow a Belgian citizen who exercised his right of free movement to stay in Belgium with his family members who previously accompanied him in another EU Member State to be unconstitutional. Consequently, the Court used three types of adjudication: (1) the annulment of the contested norm to the extent that something is not provided for; (2) the annulment of the contested norm to the extent that something is not provided for and explicitly emphasizing that it is for the legislator to fill the legislative gap; (3) a determination that the absence of legislation is unconstitutional and inciting the legislator to respond. There was a quick response to the second ruling when the law of 19 March 2014 was enacted. For the other three rulings, a circular by the Federal Government Department of Internal Affairs was for a long time the only 'legislative' response, even though the BCC had explicitly incited the legislator to react. 99 Despite the fact that this circular originated from the executive branch, it was also not sufficient to implement the rulings by the Court. Finally, by means of a Statute of 4 May 2016, the legislator implemented the rulings of the Court in an adequate manner. 100 The three judgments show that different types of dialogue unfold in highly political cases, with each time different positions for the BCC and the legislator. These different types of adjudication do not however impede redress from being offered in all cases by an expansion of the field of application. This has been done by the BCC itself when it has annulled certain words or competences, thereby immediately expanding the field of application of the contested norm. The Court has also offered ordinary courts the possibility to provide the necessary redress by granting them the competence to fill the legislative gap. Finally, in certain rulings the BCC has shown great deference to the legislator's prerogatives by highlighting that only the latter can rectify the unconstitutional omission. Eventually, the legislator filled the legislative gaps for all three judgments in accordance with the rulings of the BCC.

Improving the practice of interpreting, reviewing, writing or amending constitutions
If Meuwese and Snel refer to the actual practice, 101 i.e. the procedural and formal aspects of the dialogue, several elements were highlighted throughout this contribution that would improve the constitutional dialogue: a thorough assessment of the rulings of the Court in parliamentary proceedings, an increase of the usage of the rulings by the Court as a means of political control and, above all, the establishment of a wellfunctioning parliamentary committee responsible for monitoring the case law of the Constitutional Court.
One might assume, however, that this last aspect of the definition refers to the substantive content of a constitutional interpretation; falling back on their own expertise, the Constitutional Court and the legislator will work together to improve the protection of constitutional values. 102 When a legislative omission violates the principles of equality and non-discrimination, there are two possibilities to restore equality: either via a level-down approach, where the existing norm is removed so that no one may benefit from it any longer, or via a level-up approach, where the field of application is expanded so that all parties concerned may benefit from it. 103 Only when a higher norm obliges the legislator to provide for a certain right can the level-down approach not be followed.
Utrecht Law Review | Volume 14 | Issue 1, 2018 Decisions by the Court often imply that an enlargement of the field of application of the contested norm would resolve the unconstitutionality. Especially since ruling n° 111/2008, where the Court for the first time granted ordinary judges the competence to expand the field of application in order to fill the legislative gap, this line of reasoning is clearly present in the case law of the Court. One remark needs to be made, however: in an important lacuna judgment, the Court itself used a level-down approach to eliminate the discrimination despite the clear level-up approach which the parties requested the Court to apply. 104 In this ruling the Court stated that a tax rule was unconstitutional because it only subsidized commuter traffic for staff members of Colleges, and not for staff members of Universities. Because of the decision of the Court, it was no longer possible for the Colleges to profit from the favourable tax rule. The legislator followed this level-down approach by abolishing the subsidy for commuter traffic for staff members of Colleges.
Despite this one exception, the BCC usually strives for a level-up approach. Moreover, we have seen that if the legislator reacts, the opinion of the BCC is followed in the vast majority of legislative responses. 105 As stated before, in 72 out of the 76 legislative responses, the legislator followed the opinion of the BCC. Only in the aforementioned case did the BCC explicitly opt for a level-down approach. This means that in as many as 71 out of the 76 legislative responses (93%), the legislator (following the BCC) opted for a level-up approach, thus expanding the provided benefit or the (procedural) protection to the excluded category. What is more, after two rulings, the legislator initially opted for a level-down approach (2/76). Both rulings concerned the different tax regimes that were applicable to autonomous municipal companies and intermunicipal organizations. The Court found it discriminatory that the latter were subjected to a more favourable tax on legal persons, whereas the first were subjected to a more stringent corporate tax. The legislator initially amended the legislation so that both institutions were subject to the more stringent corporate tax. This legislative change, however, resulted in a new inequality so that after barely six months the tax legislation was once again amended, this time via a level-up approach so that both institutions could benefit from the more favourable tax on legal persons. 106 In one specific case it became very clear that the Constitutional Court and legislator work together to improve the protection of fundamental rights. 107 The case at hand has already been dealt with and pertains to the Weapons Act. As stated before, the stringent Act provided officials with unlimited access, at any time, to all locations where authorised persons conduct their activities. The BCC endorsed the legislator's idea that the goal of this far-reaching competence, namely the protection of public safety by preventing the illegal possession of firearms, could justify a deviation from the general rules regarding house searches. However, the absence of any safeguards for the rights of persons being investigated was deemed to be disproportionate to the pursued aim. Consequently, the BCC annulled this far-reaching competence for the authorities. Because of this simple annulment, the general rules regarding the inviolability of homes became applicable in cases relating to possible violations of the Weapons Act. Given that these general rules are more stringent for the officials, a better protection for the right to respect for private and family life was provided for. Although the BCC granted the legislator the possibility to provide far-reaching competences to officials if they are accompanied by certain safeguards for the protection of persons being investigated, the Minister explicitly considered that no legislative response would be initiated. He wished to adhere to the general rules regarding house searches, seeing that they offer better protection for the right to respect for private and family life. 108

Concluding remarks
Looking at the interaction between the Constitutional Court and the legislator regarding legislative omissions, elements of a constitutional dialogue clearly emerge. First, the Court often emphasizes the existence of an unconstitutional omission, sometimes explicitly referring to the need for a legislative response. However, it has been demonstrated that the phrasing of the Court's ruling, even inserting an explicit appeal to the legislator, does not necessarily incite a legislative response. Furthermore, if a legislative reaction was undertaken, in most cases the decision of the Court was at least mentioned during parliamentary proceedings. Taking into account the strict time limit which the Council of State often has, the rather limited amount of references to judgments of the Constitutional Court may not come as a surprise.
The initiatives taken by MPs proved to be less successful than bills initiated by the Government. This does not mean that MPs cannot play a crucial role within the dialogue by exercising means of political control. However, research shows that only limited use is being made of the possibility to ask parliamentary questions regarding the reaction to cases of the Constitutional Court. Unfortunately, also the Parliamentary Committee for the Evaluation of Legislation, despite its institutional design, has proved to be inadequate to exercise actual political control or to carry any weight in the legislative process. Furthermore, it became clear that references to lacuna judgments in newspaper articles are not explicitly taken into account by the legislator.
The constitutional dialogue is not restricted to the Constitutional Court and the legislature. Sometimes decrees for the executive have offered a response, and even circulars from the judiciary can be seen as a normative reaction to the judgments of the Court. When the legislator, in the broadest sense, reacts to a judgment of the Court, it almost always follows the judgment using a level-up approach; by expanding the field of application, it provides the benefit or (procedural) protection for the excluded category.
The research shows that neither the Constitutional Court nor the legislator has a dominant position within the constitutional dialogue. First and foremost, it has become clear that the legislator does not always respond to the rulings of the BCC; 31 rulings have not received a legislative reaction. What is more, in the aftermath of several cases the legislator has amended the contested norm, although the Court found that the unconstitutional omission was not situated in this contested norm. Also, in a few cases the legislator explicitly decided that, contrary to the ruling of the Court, a legislative response was not needed or a levelup approach was not necessary.
The foregoing does not however imply that the legislator can be seen as a dominant actor. Although the BCC shows great deference to the legislator's prerogatives, it sometimes changes its case law to offer ordinary judges the possibility to fill the legislative gap. For 10 out of those 31 rulings where no legislative reaction has as yet been registered, the BCC has provided the necessary redress by granting ordinary judges the competence to expand the contested norm's field of application. This possibility to offer ordinary judges the power to fill the legislative gap obliges the legislator to respond if it disagrees with the solution offered by the courts. Consequently, an important interaction takes place between the Belgian Constitutional Court and the legislator. In order for the BCC to keep fulfilling its role as a correspondent within the dialogue, it is important that ordinary judges keep referring preliminary questions to the Court. In this way, the case law of the Belgian Supreme Court (Court of Cassation) in which it declines to refer preliminary questions to the BCC when the contested norm contains an (unconstitutional) legal gap which only the legislator may fill cannot be approved. 109 Although a descriptive approach was used for this contribution, I conclude with some very preliminary normative statements on the different actors involved in the dialogue. First, the legislative section of the Council of State could be a very important player within the dialogue, but the urgency procedure, which