Introduction
More and more states, especially in Europe, have started to introduce bottom-up instruments of direct democracy in recent years and decades.Footnote 1 As the number of initiatives launched by citizens over the years demonstrates, use of these instruments is on the rise.Footnote 2
As direct-democratic participation has increased, so have concerns that initiatives and referendums might be misused, not least by populist movements, to undermine liberal democracy and oppress minorities. Citizens’ initiatives such as those to ban minaretsFootnote 3 and face coveringsFootnote 4 in Switzerland or those directed against same-sex marriage in several European statesFootnote 5 have fuelled these concerns. It was in reaction to these developments that, in 2022, the Council of Europe’s Venice Commission revised its Code of Good Practice on Referendums (Venice Commission Code).Footnote 6 Aiming to ensure that direct-democratic participation does ‘not go against international standards in the field of human rights, democracy and the rule of law’,Footnote 7 the Code – the only document setting forth international guidelines in this area – requires member states to check the ‘procedural’ and ‘substantive validity’ of texts before they are put to a referendum.Footnote 8 In their striving to balance popular sovereignty with the rule of law, European states have, accordingly, imposed various restrictions on the use of direct-democratic instruments, including by removing certain issues from the grasp of the people altogether.
Yet why would protection of the rule of law make it necessary to prevent the people from initiating certain requests? In a representative system of democracy, parliament is, as a general rule, allowed to debate, and decide upon, any issues falling within its competence. Nothing would prevent it from, say, debating and passing a law that restricts access to abortion. Measures passed by parliament can normally only be reviewed, if at all, ex post. The same holds true for those US states that provide for the citizens’ initiative: there, any proposals may be put to a referendum; their legality may normally only be challenged after the popular vote.Footnote 9 In contrast, virtually all European states with bottom-up instruments of direct democracy remove certain issues from the people’s reach from the very start: initiatives are reviewed before a popular vote may be held, often even before the collection of signatures may begin. In Spain, for example, the Constitutional Court ruled that it was not permissible to collect signatures for a citizens’ initiative that was aimed at restricting access to abortion since fundamental rights could not form the subject of initiatives.Footnote 10
Article 25(a) of the International Covenant on Civil and Political Rights (ICCPR),Footnote 11 ratified by all Council of Europe member states, guarantees every citizen the right to take part in the conduct of public affairs, without discrimination and ‘without unreasonable restrictions’. This guarantee does not oblige states to establish means of direct-democratic participation such as citizens’ initiatives. Yet where they do establish such means, the UN Human Rights Committee has held, they must refrain from imposing ‘unreasonable restrictions’ on them.Footnote 12 This raises the question of whether the admissibility requirements that are imposed on citizens’ initiatives across Europe are reasonable, and therefore justified, or ‘unreasonable’ and therefore in violation of Article 25(a) ICCPR.
To answer this question it is necessary, first of all, to establish some basic issues: What exactly are these admissibility requirements? What is the point of precluding certain types of direct-democratic requests? Who gets to decide whether a given citizens’ initiative complies with the existing admissibility requirements? Do the procedures for making this decision themselves live up to the rule-of-law standards that they are said to protect? Despite their great relevance for democratic practice across Europe, these basic issues have so far remained largely unexplored. While there are a few studies that consider some of them with a focus on specific national systems,Footnote 13 a comparison of the relevant laws and practices of all – or even only some – Council of Europe member states has not yet been undertaken. This article constitutes the first systematic attempt to categorise and critically analyse admissibility requirements and procedures for citizens’ initiatives in Europe.
The article begins with an explanation of the methodology and the categorisation of direct-democratic instruments underlying this analysis. It then explores the different types of admissibility requirements that are imposed on citizens’ initiatives, highlighting the fundamental distinction between formal and substantive requirements. This is followed by a critical assessment of the procedures used to review compliance with existing admissibility requirements and a general conclusion.
Methodology
The present comparative analysis is mainly based on a ‘functionalist’ approach.Footnote 14 It investigates what different types of admissibility requirements for citizens’ initiatives and procedures for controlling compliance with these requirements exist, what functions they serve, and whether they in fact perform these functions. In addition, the article explores the extent to which the rules of the Venice Commission Code, which are not legally binding as such, reflect state practice and may thus be regarded as common European standards. In that regard, the approach could be described as ‘universalist’.Footnote 15
In terms of its geographical scope, the analysis covers the 46 states that are members of the Council of Europe. As such they have committed themselves to the fundamental values enshrined in the Statute of the Council of Europe, including democracy and the rule of law.Footnote 16 The rule of law restrains political power by requiring legal certainty, access to justice before independent and impartial courts, respect for human rights, and equality before the law.Footnote 17 Hence, the normative premise for all European states is that they must strive to realise democracy and the rule of law at the same time, that neither of them can take absolute priority over the other. If popular sovereignty and the rule of law come into conflict, an appropriate balance must be struck between them. The article investigates how states try to strike this balance by imposing admissibility requirements on citizens’ initiatives.
The analysis builds on a large body of data that has been collected as part of the research project ‘Popular Sovereignty vs. the Rule of Law? Defining the Limits of Direct Democracy’ (LIDD), resulting in three databases that contain information on, respectively, the legal regulation of the direct-democratic instruments existing at the national level in all Council of Europe member states,Footnote 18 citizens’ initiatives launched in these states since 1990,Footnote 19 and referendums held in these states since 1990.Footnote 20 All the approximately 17,000 entries in the database concerning the regulation of direct-democratic instruments, which form the main focus of the present study, have been verified by country experts.
Given the large number of instruments compared, it is impossible to discuss the relevant practice in all countries in detail or to contextualise all findings. Furthermore, it needs to be acknowledged that there are limits to what can be gained from comparing established constitutional democracies, such as Switzerland, with regimes that are characterised by their own political leaders as ‘illiberal’, such as Hungary.Footnote 21 Nevertheless, to find out to what extent the wider constitutional and political context influences direct-democratic practice, it is, first of all, necessary to get a sense of how the respective instruments are regulated. The point of the present comparison is thus to provide a systematic overview and a first evaluation of admissibility requirements for citizens’ initiatives in Europe that hopefully will help lay the basis for future research.
Bottom-up instruments of direct democracy in Europe
In line with an established distinction in research on direct democracy,Footnote 22 the conceptual framework of the LIDD project as depicted in Figure 1 categorises direct-democratic instruments into two primary types: (1) top-down, when the process is initiated either by state institutions or automatically because certain conditions specified by law are met (such as a referendum requirement for ratifying constitutional amendments) and (2) bottom-up, when it is initiated by citizens.

Figure 1. Typology of direct-democratic instruments
Although admissibility requirements also exist for some top-down instruments, the focus of this article is on those instruments that allow voters themselves to initiate the process and shape direct-democratic requests. Within this category of bottom-up instruments (or ‘citizens’ initiatives’), a distinction is made between the proactive citizens’ initiative, the rejective citizens’ initiative, and the agenda initiative. Only instruments that exist at the national – as opposed to the local – level are covered.
The proactive citizens’ initiative allows a specified number of voters to initiate a referendum (that is, a popular vote) and formulate its topic. The proactive initiative exists in 20 out of the 46 Council of Europe member states, many of which are ‘new’ democracies: Albania, Armenia, Azerbaijan, Bulgaria, Croatia, Georgia, Hungary, Latvia, Liechtenstein, Lithuania, Moldova, Montenegro, North Macedonia, Poland, Portugal, San Marino, Serbia, Slovakia, Switzerland, and Ukraine.
The rejective citizens’ initiative allows voters to initiate a referendum that is aimed at (1) preventing constitutional or statutory provisions from being passed or (2) repealing existing ones. Nine Council of Europe states, most of them ‘old’ democracies, have the rejective initiative: Albania, Italy, Liechtenstein, Luxembourg, Malta, San Marino, Serbia, Slovenia, and Switzerland.
The agenda initiative is the weakest bottom-up instrument. It allows voters to place an issue on the agenda of a state body, typically parliament, without directly leading to a referendum. An agenda initiative that is supported by the required number of signatures obliges parliament to deal with the proposal, either by deciding on it or at least debating it. This instrument exists in 28 Council of Europe states: Albania, Andorra, Armenia, Austria, Azerbaijan, Bulgaria, Denmark, Estonia, Finland, Georgia, Greece, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Montenegro, Netherlands, North Macedonia, Poland, Portugal, Romania, San Marino, Serbia, Slovakia, Slovenia, Spain, and Ukraine. Despite not triggering a referendum, the agenda initiative is commonly classified as a direct-democratic instrumentFootnote 23 – and this for good reasons, as it can be very effective in opening up the political agenda-setting process to citizens and mobilising marginalised groups to influence parliamentary decision-making. This can be seen, for instance, in Finland, where this instrument has become a widely used channel for political participation.Footnote 24 While its direct legislative impact may often be limited, its indirect effects on agenda-setting can be significant.Footnote 25 Probably not least due to its introduction at the EU level in the form of the European Citizens’ Initiative, the agenda initiative has become very popular at the national level and is now in fact the key driver of bottom-up direct-democratic activity across Europe.Footnote 26 Since it does not automatically lead to a referendum, the Venice Commission Code may not be directly applicable to the agenda initiative. However, just like the proactive initiative and the rejective initiative, it is covered by Article 25(a) ICCPR as a means to ‘take part in the conduct of public affairs’ and may therefore not be subject to any ‘unreasonable restrictions’.
Out of the 46 Council of Europe member states, 33 have at least one bottom-up instrument of direct democracy. In total, 57 such instruments exist across Europe. Tables 1 to 3 in the Appendix give an overview of the key features of these 57 instruments with regard to admissibility requirements and procedures.
Admissibility requirements
Any use of a bottom-up instrument of direct democracy will depend on its meeting certain requirements. Here, the focus is on those requirements that determine which types of requests (that is, initiatives) are legally permissible and which ones are not. In contrast, this study does not cover requirements that are designed to measure the political support for a given initiative, that is, those concerning the collection of signatures. Of course, requirements concerning the signature threshold – varying, for the proactive initiative, from 0.56% (Portugal) to 13.5% (Armenia) of the electorate – or the collection period – ranging from 15 days (Croatia) to 540 days (Switzerland) – may have a significant impact on how frequently a given instrument is used.Footnote 27 That signature requirements are needed, however, simply follows from the nature of the citizens’ initiative; they are not requirements that are intended to prevent misuse or to protect the rule of law.
As far as requirements relating to the permissibility of initiatives are concerned, a closer look reveals that most states make a distinction between formal requirements, that is, requirements regarding the formulation of the initiative (such as the clarity requirement), and substantive requirements, that is, requirements regarding the content of the initiative (typically precluding initiatives on certain subject matters). Virtually all European states have established formal and/or substantive admissibility requirements. Only two of the 57 bottom-up instruments of direct democracy lack such requirements altogether: Luxembourg’s rejective initiative and the Austrian agenda initiative.
Formal requirements
Formal requirements for the admissibility of citizens’ initiatives are commonly understood to follow from the freedom to vote, that is, voters’ freedom to form an opinion and to give genuine expression to their will.Footnote 28 The freedom to vote is guaranteed by many national constitutionsFootnote 29 as well as by Article 25 ICCPR, which not only applies to elections but also to direct-democratic forms of participation.Footnote 30 Voters will only be able to form an opinion and express their will on a given direct-democratic request if it is formulated in a way they can understand. In that sense, formal admissibility requirements serve to put issues within the grasp of the people. They help ensure that the will of the people is formed in a way that conforms to the rule of law, central elements of which are the requirements of foreseeability and transparency.Footnote 31
Types of formal requirements
When one codes inductively the requirements relating to the formulation of the initiators’ request that exist across Europe, the following emerge as the key standards: the requirements of clarity, unity of substance, unity of form, and submission of a draft legal text (Figure 2). Except for the last one, they all appear in the Venice Commission Code.

Figure 2. Types of formal requirements
For direct-democratic requests that lead to a referendum, the Venice Commission Code defines the clarity requirement as follows:
The question put to the vote must be clear and comprehensible; it must not be misleading; it must be unbiased, not suggesting an answer; voters must be informed of the effects of the referendum.Footnote 32
This requirement is very common, especially for the proactive initiative. Article 115(6) of the Portuguese Constitution, for example, provides that initiative proposals must be formulated in an objective, clear, and precise manner and must solicit ‘yes’ or ‘no’ answers.Footnote 33 As the Supreme Court of Latvia has held, it must be unambiguously clear to any person signing – or voting on – a proactive initiative what it says, what is meant by it, and what its purpose is. Importantly, as is also made explicit by the Venice Commission Code, this implies that voters must be able to foresee the consequences of supporting a given proposal.Footnote 34 Even an initiative that is textually clear may fall short of this requirement if it defines a goal but fails to specify a mechanism for achieving it.Footnote 35
The clarity requirement is especially difficult to meet in Hungary where the Referendum Act states that clarity needs to be ensured not only for the voters but also for the legislator.Footnote 36 Thus, parliament must be able to tell from the wording of the proposal how it is expected to implement it.Footnote 37 These two aspects of clarity may often be in tension, as a very precisely formulated proposal may be easier for parliament to implement but more difficult for voters to understand.Footnote 38 The result of these demanding standards is that, from 2014 to 2017, 68% of proactive initiatives in Hungary were declared inadmissible due to a lack of clarity.Footnote 39 For example, the Curia (the Hungarian Supreme Court) held that an initiative demanding that all shops should be closed on Sundays was not sufficiently clear. It argued that, given how the initiative would have to be implemented by parliament, voters would not be able to foresee all the consequences its approval would bring about to everyday life.Footnote 40 In several countries, the clarity requirement also applies to the rejective initiative. It has acquired especially great importance for the Italian version of this instrument, the referendum abrogativo. Despite the lack of a respective explicit basis in Italian legislation, the Italian Constitutional Court has consistently held that, to be admissible, abrogative requests must be simple, unambiguous, and complete.Footnote 41 For instance, it ruled that a rejective initiative aimed at abolishing the electoral legislation in force at the time failed to meet this requirement as it did not allow voters to foresee what kind of electoral regime would replace the existing one.Footnote 42
According to the unity of substance requirement, which is normally regarded as following from the clarity requirement,Footnote 43 a citizens’ initiative must not combine two or more issues or subject matters that lack an inherent connection.Footnote 44 If the requirement is not respected, signatories and voters cannot freely express their will since they may be in favour of one aspect of the initiative but opposed to another. The requirement does not apply to the total revision of a constitution or law. Unity of substance is an explicit requirement for roughly half of the proactive initiatives and rejective initiatives existing in Europe. For the Ukrainian proactive initiative, for example, the relevant law provides that, if a referendum proposal consists of several parts, there must be an inseparable internal connection between them.Footnote 45 In Switzerland, unity of substance is required by Article 139(3) of the Federal Constitution. It is on this basis that the Federal Assembly declared inadmissible an initiative that demanded that the military budget be reduced by half while at the same time requiring that this money be used for purposes of peacekeeping and the provision of social security.Footnote 46 Unity of substance equally plays an important role for the Italian rejective initiative, as this instrument also allows initiators to ask for a repeal of several legal provisions, which need not even form part of the same law. In a case concerning a request aimed at the – total or partial – repeal of 97 articles of the penal code, the Italian Constitutional Court held that a rejective initiative may not include a plurality of heterogeneous questions that lack a rationally uniform aim. If there is a lack of unity of substance, the Court argued, the people cannot express their will genuinely, which violates their freedom to vote and runs counter to the very concept of the referendum.Footnote 47 Since then, the Court has consistently subjected abrogative requests to a strict review of their compliance with the unity of substance requirement.Footnote 48
The unity of form requirement means, in the words of the Venice Commission Code, that an initiative ‘must not combine a specifically worded draft amendment with a generally worded proposal or a question of principle’.Footnote 49 Whereas a specifically worded amendment can be enacted as it is, approval of a generally worded proposal or a question of principle is simply a first step in the law-making process that must be followed by other steps. Therefore, if the requirement is not complied with, voters cannot foresee the exact consequences of a ‘yes’ vote. In that sense, unity of form can also be regarded as a specific aspect of clarity. The unity of form requirement only applies to five proactive initiatives, three rejective initiatives, and one agenda initiative. Where it does exist, it is of little practical relevance.Footnote 50
Many states set higher standards than those established by the Venice Commission Code and require initiators to submit their proposal in the form of a draft legal text, meaning that generally worded proposals are not allowed. In Romania, for example, agenda initiatives need to be submitted in the form that is required for draft legislation, accompanied by an explanatory memorandum.Footnote 51 This requirement may also serve to protect the freedom to vote: it will generally be easier for signatories and voters to understand a direct-democratic request if it is formulated as a detailed legal text rather than a general proposal. At the same time, however, the requirement also facilitates the work of parliament as it will not itself have to come up with the legal provisions that, in the case of the proactive initiative and the rejective initiative, will be submitted to the referendum or, in the case of the agenda initiative, will form the basis of parliamentary debates. At least in the case of most agenda initiatives, this appears to be the main objective of the draft text requirement. If the point was to ensure signatories’ freedom to form an opinion, compliance with the requirement would have to be checked prior to the signature collection, which only seven of the 17 states that have this requirement for their agenda initiatives do.Footnote 52 Like other formal requirements, the draft text requirement may amount to a high hurdle for initiators, especially less politically active ones. The Latvian Central Election Commission, for instance, tends to apply restrictively the requirement that initiators must ‘submit a fully elaborated draft of an amendment to the Constitution or of a law’Footnote 53 and has declared several proactive initiatives inadmissible on this basis.Footnote 54
Other formal requirements include, for example, that with a proactive initiative not more than oneFootnote 55 or threeFootnote 56 questions may be submitted to a referendum; that the question proposed by a proactive initiative may not be expressed in such a way as to give priority to, or suggest, one of the possible answers;Footnote 57 or that the wording of a rejective initiative must follow a set format determined by law.Footnote 58
Diffusion
Since most requirements regarding the formulation of direct-democratic requests follow from the freedom to vote, they are very common across Europe. They normally apply not only to bottom-up but also to top-down instruments.Footnote 59 They are more common than substantive requirements. Only eight out of the 57 bottom-up instruments existing in Europe are not subject to any formal requirements (Figure 3).

Figure 3. Instruments with formal requirements
All 20 states that have the proactive initiative impose at least one formal requirement on its use, with the clarity requirement (16 states) and the unity of substance requirement (9 states) being the most common ones. This makes sense, as the proactive initiative allows the initiators to formulate their own proposal. This proposal not only forms the basis for the collection of signatures, it is – if the required number of signatures is collected – then also put to a popular vote. It is crucial that at both these stages – the signature collection and the referendum – voters have a clear understanding of what exactly is requested.
In the case of the agenda initiative, the initiators’ proposal equally forms the starting point for the signature collection. However, if sufficient signatures are collected, the initiative is then submitted to parliament rather than the people. In most states, parliament enjoys wide discretion in how it deals with the proposal and may thus also adapt its text at will. This explains why only a small minority of states have requirements of clarity (six out of 28 states) or unity of substance (four out of 28 states) for the agenda initiative. In contrast, in 17 out of 28 states an agenda initiative needs to be submitted as a draft legal text.
The situation is somewhat different for the rejective initiative. Like the proactive initiative, it leads to a referendum. It is therefore of crucial importance that both signatories and voters can freely form an opinion on the text at issue. However, since the initiators do not draft their own proposal but oppose norms that are proposed (or have already been adopted) by parliament, they have less influence on the text voted on. In Slovenia and Malta, for example, the initiators’ request must follow a set format along the lines of ‘Should the following Act/provisions enter into/stay in force?’.Footnote 60 Nevertheless, especially where, as in Italy and San Marino, the rejective initiative may be used to abrogate single provisions (or even only parts of provisions) of existing laws – and thereby also to change their meaning – there is a risk that such requests may be unclear or even misleading. Therefore, six out of nine states have either the clarity or the unity of substance requirement (or both) for their rejective initiative.
Assessment
For an issue to come within voters’ reach, it must meet certain standards regarding its formulation. Formal admissibility requirements such as clarity and unity of substance can thus be said to follow from the very idea of direct democracy: the people can only genuinely express their will regarding a direct-democratic proposal if it is clear and homogeneous. This fundamental connection has been highlighted, for example, by the Italian Constitutional Court and the Latvian Supreme Court.Footnote 61
Given their fundamental importance, it is problematic that for one rejective initiative and seven agenda initiatives the relevant national laws do not provide for any formal admissibility requirements at all. For agenda initiatives, formal requirements may seem less vital than for the other instruments, considering that they do not lead to a popular vote. Yet not only voters but also potential signatories should have a clear understanding of initiative proposals.
While true direct-democratic participation depends on formal admissibility requirements, these may also be used to undermine it. As the Hungarian example demonstrates, the clarity requirement can turn into a serious hurdle for citizens’ initiatives, especially if it is assessed primarily from the perspective of the authorities rather than that of voters. Similarly, the requirement of submission of a draft legal text may serve to protect the freedom of vote but, depending on how it is applied, it may also become an obstacle for direct-democratic participation as voters may struggle to meet demanding editorial standards.
Substantive requirements
The vast majority of European states allow citizens’ initiatives to be declared inadmissible because of their content and hence exclude certain issues from decision-making by the people. The Venice Commission Code requires proposals submitted to a referendum to ‘comply with all superior law’ and not to ‘be contrary to international law, to the Council of Europe’s statutory principles (democracy, human rights and the rule of law) or to Council of Europe membership conditions’. It then adds that ‘[s]tates may add further limitations’.Footnote 62 Many states have done so: they also preclude initiatives that touch on certain subject matters, fail to raise issues of special importance, or would require an amendment of the constitution.
Types of substantive requirements
The degree of content-related restrictions that are imposed on bottom-up instruments of direct democracy varies considerably from state to state. Whereas the Portuguese Constitution precludes more than 30 subject matters from the scope of the proactive initiative,Footnote 63 the only substantive admissibility requirement that exists in Switzerland for the same instrument is conformity with peremptory norms of international law.Footnote 64 Nevertheless, an analysis of all 57 bottom-up instruments existing in Europe reveals a set of characteristic substantive requirements. Figure 4 lists the most common ones.

Figure 4. Prohibited subject matters
There is no obvious way of categorising these substantive requirements, as a given requirement – such as compliance with international law – may serve several objectives at the same time and touch upon various subject areas.
One obvious rationale behind some of the substantive requirements – in particular those of compliance with international law, the constitution, fundamental rights, and minority rights as well as respect for the very structure of the state – is the attempt to balance popular sovereignty with the demands of the rule of law. Norms that are regarded as fundamental for a democracy based on the rule of law are removed from the reach of the people. However, as pointed out in the Introduction above, compliance of citizens’ initiatives with such norms could also be ensured at a later stage. Why, instead, is it made a requirement for their admissibility? Apparently, the intention is to prevent a dilemma from arising: if a proactive initiative or a rejective initiative that clashes with fundamental legal norms is allowed to go ahead and is approved in a popular vote, then there is only the choice to either respect the voters’ will and thus disregard these legal obligations or to respect these obligations and thus disregard, or at least not fully implement, the voters’ will. The first solution creates legal uncertainty, the second is bound to lead to political controversy. A similar dilemma may arise in the case of an agenda initiative that is supported by a large number of signatories. As the Slovenian Constitutional Court has pointed out, allowing signatures to be collected for an initiative that clashes with the fundamentals of a constitutional democracy might lead to political polarisation.Footnote 65
Therefore, numerous states preclude citizens’ initiatives that touch upon international law altogether. In Hungary, for example, proactive initiatives may not affect ‘any obligation arising from international treaties’.Footnote 66 The Curia declared an initiative demanding that foreign nationals be prohibited from acquiring agricultural land inadmissible on this basis as it was incompatible with Hungary’s Treaty of Accession to the European Union.Footnote 67 In several states, the requirement of compliance with international law equally applies to the rejective initiative. According to Article 75(2) of the Italian Constitution, for example, abrogative referendums may not be directed against ‘laws authorizing the ratification of international treaties’. The Constitutional Court has interpreted this subject matter prohibition widely to also cover laws that are necessary to implement obligations arising from international agreements.Footnote 68 Finally, to give an example of an agenda initiative, under Estonian law such initiatives are only admissible if they are not ‘manifestly incompatible … with the international obligations of the Republic of Estonia under international agreements’.Footnote 69 The problem with such requirements, which simply refer to ‘compliance with international law’, is that the hurdle may be set higher than is needed, given that proposals may be filtered out despite there not being a real dilemma. Should it really be the case that, say, an important constitutional amendment is barred because it may be incompatible with a bilateral agreement on a minor technical issue that is of hardly any importance for either of the states?
The example of Switzerland, on the other hand, shows that lowering the hurdle too much may also be problematic. Article 139(3) of the Swiss Federal Constitution only allows proactive initiatives to be declared inadmissible if they violate ‘peremptory norms of international law’,Footnote 70 such as the prohibitions of genocide, slavery, and torture. As a result, numerous initiatives that comply with peremptory norms but violate other norms of international law have been put to a popular vote. Some of them have been approved by a majority of voters and cantons. These include an initiative ‘against mass immigration’, which conflicts with the Agreement on the Free Movement of Persons concluded between Switzerland and the European Union,Footnote 71 and an initiative for the automatic expulsion of foreign nationals convicted of certain criminal offences, which violates several human rights guarantees.Footnote 72 The question of whether, and how, these initiatives should be implemented has led to major controversies and a veritable crisis of the political system.Footnote 73
In some states, it is not permissible to use bottom-up instruments of direct democracy to amend the constitution. This is a common admissibility requirement for the agenda initiative, existing in nine states, including SpainFootnote 74 and Denmark.Footnote 75 In three states the rejective initiative may not be used to prevent or repeal constitutional amendments. The Italian Constitutional Court, for example, has excluded from the scope of the abrogative referendum norms that have constitutional significance or that render a constitutional norm inoperable, even though the wording of Article 75 of the Italian Constitution does not set forth such a limit.Footnote 76 In contrast, in the case of the proactive initiative, a prohibition on amending the constitution only exists in HungaryFootnote 77 and Portugal.Footnote 78
Where the constitution can be amended through direct-democratic means, there may nevertheless be a requirement that citizens’ initiatives must not violate its most fundamental principles or the very structure of the state. In Estonia, for example, agenda initiatives may not be ‘manifestly incompatible with the fundamental principles of the Constitution’.Footnote 79 In Romania they may not be aimed at amending those constitutional provisions that concern ‘the national, independent, unitary and indivisible character of the Romanian State, the republican form of government’, or ‘political pluralism’.Footnote 80 Although there is no similar provision in Croatian law, the Croatian Constitutional Court has held that a proactive initiative may be declared inadmissible if ‘it threatens to destroy the structural characteristics of the Croatian constitutional state’.Footnote 81 Similarly, the Latvian Constitutional Court has held that proactive initiatives may not be ‘contrary to the fundamental values of a democratic state governed by the rule of law’, adding that, if such initiatives were regularly submitted to a popular vote, ‘the very idea of a citizens’ initiative would be undermined’.Footnote 82
A further very common substantive requirement for all types of bottom-up instruments is respect for fundamental and minority rights. In some cases, including for example that of the proactive initiative in Slovakia, the requirement is explicitly contained in the constitution.Footnote 83 In others, such as that of the Hungarian proactive initiative, it follows from the prohibition of constitutional amendments.Footnote 84 In still others, for instance that of the Croatian proactive initiative, it may be implied in the requirement to respect the structural characteristics of the state. In Slovenia, rejective initiatives may not be directed against ‘laws eliminating an unconstitutionality in the field of human rights and fundamental freedoms or any other unconstitutionality’.Footnote 85 In Romania, agenda initiatives may not result ‘in the suppression of the citizens’ fundamental rights and freedoms, or of the safeguards thereof’.Footnote 86
The scope of such prohibitions may be controversial, especially where they are as vaguely defined as Article 93(3) of the Slovakian Constitution, which simply refers to ‘issues of fundamental rights’. Do they preclude only citizens’ initiatives that violate fundamental rights, or also those that restrict them, or even all those that concern them? The Slovakian Constitutional Court has made it clear that the constitutional prohibition does not apply to initiatives that would extend the standard of protection offered by a given fundamental right.Footnote 87 Yet when exactly, then, does an initiative have to be qualified as being contrary to such a substantive requirement? This question has been raised in several cases concerning direct-democratic proposals to ban same-sex marriage. In 2013, the Croatian Constitutional Court held that a proactive initiative could only be declared inadmissible if the unconstitutionality of the proposal was ‘of such severity that it threatens to destroy the structural characteristics of the Croatian constitutional state’.Footnote 88 According to the Court, a ban on same-sex marriage as proposed by the initiative in question did not fall into this category; it was, on the contrary, ‘aligned with the European legal standards regarding the institutions of marriage and family life’.Footnote 89 Similarly, the Slovakian Constitutional Court concluded in 2014 that prohibiting same-sex couples from marrying and adopting children would not lower the standard of protection of the right to private and family life to an extent that threatens the rule of law and hence was compatible with the prohibition of Article 93(3) of the Constitution.Footnote 90 The Slovenian Constitutional Court ruled in 2015 that a rejective initiative directed against an act of parliament designed to introduce same-sex marriage was admissible as the act did not directly ‘eliminat[e] an unconstitutionality in the field of human rights and fundamental freedoms’ in the sense of Article 90(2) of the Slovenian Constitution.Footnote 91 Finally, in 2016 the Romanian Constitutional Court ruled that an agenda initiative that demanded the term ‘marriage’ in the Constitution to be limited to a union between a man and a woman did not amount to a suppression of fundamental rights as it merely specified the scope of the right to marry.Footnote 92
Conversely, initiative proposals that have been declared inadmissible due to a lack of conformity with fundamental rights guarantees include the Spanish agenda initiative aimed at restricting access to abortion referred to in the Introduction;Footnote 93 a Croatian proactive initiative proposing to restrict the use of minority languages;Footnote 94 a Slovenian rejective initiative directed against an act that provided redress for the human rights violations caused by erasure from the Slovenian register of permanent residents;Footnote 95 and, in Hungary, a proactive initiative intended to oblige the relatives of officeholders to publicly declare their assetsFootnote 96 and several proactive initiatives aimed at the reintroduction of the death penalty.Footnote 97
While all the requirements described above may be said to protect the rule of law, there are also numerous content-related restrictions that obviously serve completely different objectives.
This is true for the most common subject matter prohibition: that of financial issues, which applies to 29 instruments. In Poland, proactive initiatives may not concern ‘expenditure and revenue, especially taxes and other public duties’;Footnote 98 in Latvia, the prohibition covers ‘[t]he budget and laws concerning loans, taxes, customs duties [and] railroad tariffs’.Footnote 99 Rejective initiatives may not target ‘tax or budget laws’ in ItalyFootnote 100 or ‘laws on taxes, customs duties, and other compulsory charges, and on the law adopted for the implementation of the state budget’ in Slovenia.Footnote 101 In the Netherlands, ‘taxes and budgets’ are off limits to the agenda initiative,Footnote 102 in Spain subjects that are ‘of a tax nature’.Footnote 103 Slovenia introduced its prohibition of financial issues to ensure ‘the financial stability of the country’.Footnote 104 The similarly worded prohibition in Slovakia has been described as ‘understandable in regard to the preservation of a stable and predictable state income’ and has remained uncontroversial.Footnote 105 There thus appears to be a widely held assumption that voters cannot be entrusted with decisions that may affect the state’s financial stability as ‘it is regarded as unlikely that the people would place the overall state interest above their own’.Footnote 106
Just like substantive requirements relating to international law or fundamental rights, those concerning financial issues raise problems of interpretation. Almost any issue that can form the subject of a citizens’ initiative will have some financial implications. The Latvian Central Election Commission, for instance, has therefore interpreted the prohibition of financial issues narrowly.Footnote 107 In contrast, the Hungarian authorities have declared a large number of proactive initiatives inadmissible because they concerned fiscal legislation.Footnote 108 And the Italian Constitutional Court has held that the reference to ‘tax or budget laws’ of Article 75(2) of the Constitution also covers measures that are closely connected to budget laws, including laws that are deemed necessary for the implementation of financial and budgetary goals.Footnote 109
Quite a few states have put national security matters and/or emergency powers beyond the reach of bottom-up instruments of direct democracy. In Hungary, ‘the declaration of a state of war, state of national crisis or state of emergency’ and ‘any matter related to participation in military operations’ may not form part of a proactive initiative.Footnote 110 In a similar vein, the respective Polish prohibition covers all matters concerning ‘state defence’.Footnote 111 These substantive admissibility requirements seem to be prompted by similar considerations as those that explain exemptions of financial issues, namely a reluctance to entrust questions that may affect the stability or even the very existence of the state to the judgment of the people. When, in 1922, the subject matter prohibitions (including financial issues, declaration of war or a state of emergency, military conscription, and mobilisation) were introduced in Latvia, the expectation was that, ‘as culture spreads and civic consciousness strengthens’, they could be gradually removed.Footnote 112 In fact, however, they have been left in place until today. Another justification for putting decisions regarding national security beyond the reach of the people may be that they often require swift action. Direct-democratic procedures, which are normally time-consuming, may therefore be regarded as inappropriate means of decision-making. Accordingly, in Slovenia, only those measures to ensure the defence of the state that are urgent are removed from the reach of the rejective initiative.Footnote 113 In 2015 the Slovenian Constitutional Court ruled that, in the context of the European migrant crisis, a law allowing members of the armed forces to be deployed to control the national border constituted an urgently needed security measure, so that a rejective initiative trying to prevent its entry into force was inadmissible.Footnote 114
Diffusion
Whereas only about half of the top-down instruments of direct democracy include one or more substantive requirements,Footnote 115 they are very common for citizens’ initiatives.
Content-related restrictions exist for 46 of the 57 instruments (Figure 5). They are particularly common for the proactive initiative, which allows voters to come up with their own proposal that is then submitted to a referendum. Montenegro, where the proactive initiative appears to be of no practical relevance, is the only state that does not provide for any substantive requirements for this instrument.

Figure 5. Instruments with substantive requirements
Given that the rejective initiative serves to oppose a proposal or measure that stems from parliament, it may seem surprising that seven of the states with this instrument impose content-related restrictions on it. The point of most of these restrictions is to prevent the blocking, or repeal, of provisions that are regarded as essential for the functioning of the state, such as laws on taxes,Footnote 116 the budget,Footnote 117 or urgent measures to ensure the defence of the state.Footnote 118
As with formal requirements, substantive requirements are least common for the agenda initiative. Nevertheless, 20 out of 28 states do preclude agenda initiatives on certain subject matters, ranging from human rights,Footnote 119 to taxes,Footnote 120 to amnesties and pardons.Footnote 121 This is remarkable, considering that parliament normally has broad leeway in dealing with agenda initiatives. If it regards a given proposal as problematic in terms of its content, it could address the problem at the stage of implementation or simply reject the proposal. Thus, the point of these substantive requirements seems to be to prevent certain issues from being put on the political agenda in the first place.
To get a better sense of how high the bar is set for the various bottom-up instruments, the list of 20 prohibited subject matters (Figure 4) has been transformed into a four-point scale of degrees of substantive requirements, which takes account of the number and stringency of content-related restrictions existing for a given instrument. The result is depicted in Figure 6.

Figure 6. Degree of substantive requirements
This scale confirms the picture we get from the simple count of instruments with substantive requirements (Figure 5): the restrictions imposed on the content of proposals are particularly stringent in the case of the proactive initiative. While the requirements for the agenda initiative are comparatively relaxed, there is still a remarkably large number of states that have a medium or even high degree of substantive requirements for this rather weak instrument.
Assessment
Some of the substantive admissibility requirements that have been imposed on citizens’ initiatives, such as those relating to finances or national security, are evidently prompted by a distrust of the judgment of the people. Yet why should, in a democracy, key decisions on issues that may affect the stability of the state be reserved for members of parliament or even unelected government officials? There is no convincing reason why the people should be trusted to elect representatives, but not to decide on policies. Parliamentarians are not in a better position than voters to make responsible decisions: a large body of empirical work demonstrates that those participating in popular votes do have the necessary competence to make policy decisions,Footnote 122 and parliamentarians or government officials face exactly the same conflict of interests as voters when they decide on, say, reducing taxes or increasing public spending. In fact, the available empirical evidence suggests that direct democracy leads to a reduction in government spending.Footnote 123 Similarly, why should citizens not have a say on whether the armed forces should be deployed against migrants? Of course, such decisions might need to be taken urgently. Yet this is no reason to not involve the people in them at all. In Switzerland, for example, parliament may, in case of urgency, enact a law immediately.Footnote 124 However, voters are still allowed to launch a rejective initiative against such an emergency law. If they are successful in collecting sufficient signatures, a referendum on the law is then held within one year from enactment.Footnote 125 This shows that prohibitions of subject matters such as finances or national security are not supported by compelling reasons but mainly reflect a paternalistic attitude. Consequently, they cannot be regarded as ‘reasonable restrictions’ of direct-democratic participation in the sense of Article 25(a) ICCPR.
Content-related requirements also seem difficult to justify where they are imposed on agenda initiatives. Given the weak nature of the agenda initiative and the wide scope of discretion parliament typically has in dealing with this type of initiatives, such requirements mainly seem to serve to prevent certain subjects from being put on the political agenda.
In contrast, most lawyers will have sympathy for the Venice Commission’s recommendation that direct-democratic proposals should comply with higher-ranking law, international law, and human rights.Footnote 126 After all, democracy without the rule of law is unthinkable. A democratic system depends on legal safeguards that prevent the concentration and perpetuation of political power; those in the minority need protection to have the chance to become the political majority.Footnote 127 Seen from this perspective, the Latvian Constitutional Court was right to state that citizens’ initiatives that clash with fundamental legal norms may undermine the very idea of direct democracy.Footnote 128 This explains why many states follow the Venice Commission’s recommendation and try to prevent such initiatives from the very beginning. Otherwise, the political pressure these may generate if they collect the required number of signatures, or are even approved in a popular vote, can put the authorities in a difficult situation. However, in practice, it has proven extremely challenging to filter out the right initiatives as it is not obvious which types of norms should be off limits to bottom-up instruments of direct democracy. Definitions that simply refer to ‘international law’ or ‘fundamental rights’ risk being over-inclusive, while those that refer to ‘fundamental principles’ or ‘structural characteristics of the constitution’ raise serious problems of interpretation. All these broadly defined requirements leave those applying them wide discretion.
Admissibility procedures
As shown above, admissibility requirements are difficult to define and interpret. Hence, the questions as to who is tasked with reviewing whether citizens’ initiatives comply with them and how the process of review is designed, assume particular significance. Do these admissibility procedures live up to the demands of the rule of law? To answer this question, we first look at the institutions in charge of the admissibility review, asking whether the respective decision-makers are impartial. This is followed by an analysis of the procedural safeguards that apply: do initiators or voters have a right to be heard before the decision is taken? Can the decision be appealed? Finally, we evaluate the relevance of the timing of the admissibility check for the rule of law.
Where formal and/or substantive admissibility requirements exist for citizens’ initiatives, compliance with them is normally reviewed in a devoted admissibility procedure, meaning that a given initiative is only allowed to proceed to the signature collection or the popular vote once compliance has been confirmed. However, there are certain instruments that are subject to admissibility requirements without there being an institution that would be charged with reviewing compliance with these requirements. In these cases, initiatives are not systematically reviewed for their admissibility. Nevertheless, it is still possible that the question of compliance with formal and/or substantive requirements is raised: it may be at the discretion of an institution involved in the process to apply these requirements or a voter may contest the admissibility of an initiative before a court. Hence, an instrument might be subject to admissibility requirements without there being an admissibility procedure. In contrast, it is impossible for there to be a check without requirements: for there to be a review, there must be a threshold to be met.
Procedures for reviewing compliance with formal and substantive admissibility requirements do not necessarily coincide. Some instruments are subject to a formal but not a substantive check, others to a substantive but not a formal check. Whilst for most instruments the formal and substantive checks are entrusted to the same institution, for some instruments the institution depends on the nature of the check. Nine instruments are subject to neither a formal nor a substantive check: the proactive initiative in Montenegro, the rejective initiatives in Luxembourg and Switzerland, and the agenda initiatives in Albania, Armenia, Austria, Bulgaria, Montenegro, and San Marino.
Institution
The LIDD database on the legal regulation of direct-democratic instruments distinguishes five options regarding the institution that is tasked with reviewing the admissibility of citizens’ initiatives: there may be no admissibility check at all, or parliament (or a parliamentary committee), the government, an election commission,Footnote 129 or a court may be in charge of it.
Formal check
About a fifth of the proactive initiatives and rejective initiatives and 12 out of the 28 agenda initiatives are not subject to a formal check (Figure 7). For eight out of the 18 instruments without a check there are no formal admissibility requirements in the first place. This still leaves ten instruments for which formal requirements exist but compliance with them is not systematically reviewed.

Figure 7. Institution in charge of the formal check
Where agenda initiatives are checked for their formal admissibility, it is typically parliament that is entrusted with this task. While parliaments also play their part in the formal admissibility review of proactive initiatives, the most common institutional choice for this instrument is election commissions. Courts assume a prominent role in checking rejective initiatives, but less so in the case of the other instruments. Only in the case of four instruments is the government tasked with the formal admissibility review.
Substantive check
Some 17 instruments, 13 of which are agenda initiatives, are not subject to a substantive admissibility check (Figure 8). For 11 out of these 17 instruments no substantive requirements exist, so that a corresponding check is conceptually impossible.

Figure 8. Institution in charge of the substantive check
In the case of 25 instruments, the substantive admissibility check is entrusted to parliament. Courts equally assume a prominent role when it comes to substantive requirements, especially in the case of the proactive initiative and the rejective initiative. In contrast, election commissions and governmental bodies are a rare institutional choice for the substantive check.
Assessment
The main purpose of formal admissibility checks is to ensure that direct-democratic requests are in fact within the grasp of voters: that they can understand what a given request is about and what consequences it will entail so that they can freely form an opinion before deciding whether to support it. Seen in this light, it is highly problematic that for ten instruments that are subject to formal requirements there is no institutional and procedural framework for reviewing initiatives for their compliance with these requirements. This effectively means that not only for the eight instruments that are not subject to formal requirements in the first place, but also for these ten instruments there is no guarantee that voters will be free to form an opinion.
In contrast, substantive requirements do not serve to guarantee the freedom to vote but rather to remove certain issues from the reach of the people. At least from the perspective of voters, it may therefore appear more acceptable than in the case of formal requirements that compliance with substantive requirements is not reviewed systematically but only on request. Nevertheless, if substantive admissibility requirements are thought to be needed, it would only be logically consistent to also establish a procedure for ensuring compliance with them.
Insofar as procedures for reviewing the formal and/or substantive admissibility of citizens’ initiatives do exist, the question arises whether the institutions that are put in charge of them are suited for this task. Even though citizens’ initiatives are used to promote or oppose a (legislative) proposal, the point of the admissibility procedure is not to review the desirability or feasibility of that proposal but rather its conformity with the admissibility requirements as they are defined in law. Thus, whether a given initiative meets or does not meet these requirements – be they of a formal or a substantive nature – is a legal question. As such, it should be decided by a body with legal expertise. Institutions that routinely apply and interpret legal norms and decide legal disputes, such as courts or election commissions, are thus considerably better equipped to review the admissibility of citizens’ initiatives than parliaments, which primarily deal with political and policy questions.
Furthermore, institutions tasked with this review must be impartial, meaning that they should treat those involved in the procedure equally and reach a decision that is free of bias or prejudice. Impartiality implies that the respective institution is independent, allowing it to decide without any improper influence from outside. Article 25(b) ICCPR requires states to establish an independent authority to supervise popular vote processes and ensure that they are conducted fairly and impartially.Footnote 130 The Venice Commission Code states that an impartial body must be entrusted with the organisation of referendums, including those triggered by citizens’ initiatives.Footnote 131 That impartial body must also have the power ‘to check the validity of any proposed referendum question and approve its final wording’.Footnote 132
Courts are ideally placed to perform this task. In most European states, a variety of institutional features protect them from undue outside pressures. As highlighted by the Venice Commission, the admissibility check may also be entrusted to election commissions as long as their independence is guaranteed, implying that they should include at least one independent legal expertFootnote 133 and that their members may not be dismissed at will.Footnote 134
The impartiality of parliaments, in contrast, may be compromised for two basic reasons. First, members of parliament may take into account strategic considerations when reviewing citizens’ initiatives. Being dependent on electors’ support, their decision-making may be influenced by the popularity of a given initiative: They may be reluctant to declare initiatives inadmissible that appear to enjoy wide support among their voters. Conversely, they may feel urged to vote for the inadmissibility of unpopular proposals. Second, ideological considerations may also influence decision-making in parliament. Given that citizens’ initiatives are intended to remove certain issues from the policy-making authority of parliament, its members may have an interest in thwarting initiatives that do not align with their political preferences.Footnote 135 Similar strategic and ideological considerations may also affect the decision-making of governmental bodies, although – depending on the type of body – its members will normally have to vie less for voters’ favour than parliamentarians.
Considering the real risk that the impartiality of parliaments and governmental bodies may be compromised by strategic and ideological considerations, it is of great concern that a relatively large number of states entrust either type of these institutions with reviewing citizens’ initiatives for their admissibility. This concern can be alleviated by providing for the possibility to appeal parliamentary or governmental decisions before a court. However, as will be explained in the following section, such a legal remedy does not exist in all states.
Procedural safeguards
To what extent do procedures for reviewing the admissibility of citizens’ initiatives incorporate safeguards for initiators or voters? The LIDD database contains entries on, first, the right of initiators and voters to participate in these procedures, be it in the form of an oral hearing before the decision-maker (‘right to be heard’) or through written submissions. Second, it includes information on the remedies that are available to challenge decisions declaring initiatives inadmissible: is there a right to appeal such decisions? If so, who is allowed to bring an appeal? Which type of institution decides the appeal?
The rights to participate and to appeal are defining elements of a procedural understanding of the rule of law. They both allow interested parties to control how the decision-maker fulfils its task, thus enhancing the transparency of the admissibility procedure. Furthermore, the right to participate helps ensure that all information required for reaching a well-grounded decision is at hand, and the availability of an effective remedy may help prevent or correct decisions that are ill-founded or even arbitrary. Finally, these rights serve to ensure that the interested parties are not treated as objects of the procedure but as subjects who deserve respect.Footnote 136
Formal check
Procedures for reviewing the formal admissibility of citizens’ initiatives only rarely incorporate a right to participate. Such a right only applies to 13 out of the 39 instruments for which there is a formal check (Figure 9). The right exists, for example, in North Macedonia where the parliamentary committee tasked with dealing with a given agenda initiative must notify the initiators’ representative of any relevant sessions and allow him or her to participate in them.Footnote 137

Figure 9. Procedural safeguards: formal check
The right to an effective remedy is more common: 19 instruments allow for the possibility of appealing the first-instance decision declaring an initiative inadmissible on formal grounds. In 18 of these cases, it is a court that will decide the appeal. The only exception concerns the agenda initiative in Portugal where inadmissibility decisions by the president of the parliament may be appealed before parliament as a whole.Footnote 138 In the case of rejective initiatives and agenda initiatives the right to appeal is almost exclusively granted to the initiators, whereas in the case of proactive initiatives several states allow any voter to appeal (in)admissibility decisions.
Substantive check
Procedural safeguards are similarly rare in the case of substantive checks of citizens’ initiatives. 15 out of the 40 instruments for which there is a substantive review allow for a right to participate (Figure 10). The right to appeal decisions declaring an initiative substantively inadmissible is available for 13 instruments. As with the formal check, it is, except for the Portuguese agenda initiative, courts that decide on appeal, and the right to challenge first-instance decisions is normally restricted to the initiators.

Figure 10. Procedural safeguards: substantive check
Taking together remedies available in formal and substantive admissibility checks, decisions by election commissions can be appealed in nearly all cases, those by parliament or government in roughly half of the cases. In contrast, where a court decides on admissibility in the first instance, that decision cannot be appealed in any state.
Assessment
The Venice Commission Code lists numerous procedural guarantees that apply to the organisation of referendums, including those triggered by a proactive initiative or a rejective initiative. However, it should be noted that these guarantees have been designed generically for all parts of the referendum procedure, from regulation of the franchise to campaigning issues and the establishment of referendum results; they are not specifically tailored to admissibility checks.
The Venice Commission Code refers to the right to participate in the context of the system of appeal that states need to install in referendum matters, requiring the applicant’s right to an adversarial hearing to be protected.Footnote 139 Considering its importance in terms of ensuring the availability of all necessary information and enhancing transparency, it is of concern that only a small minority of existing admissibility checks provide for the right to participate. Both when it comes to formal and substantive admissibility, a thorough assessment may well depend on information that the initiators – or other actors with a special interest or expertise in the subject matter at hand – are best able to provide.
Article 25 in conjunction with Article 2(3) ICCPR guarantees voters access to an effective remedy in case of disputes concerning citizens’ initiatives.Footnote 140 Whether Article 14(1) ICCPR, providing a right of access to a court, also applies to such disputes, has not yet been settled. The UN Human Rights Committee has applied Article 14(1) ICCPR to electoral disputes,Footnote 141 which suggests that it could be equally applicable to disputes concerning the admissibility of citizens’ initiatives. If access to judicial review is not available, there must, at the very least, be access to an ‘equivalent process’.Footnote 142 In a similar vein, the Venice Commission Code requires states to establish an ‘effective system of appeal’ in referendum matters.Footnote 143 The appeal body, which must also have the competence to address the formal and substantive admissibility of citizens’ initiatives,Footnote 144 should be impartial and independent and endowed with the necessary powers to afford an effective remedy. The Code stresses that a final appeal to a court of law is the preferred option.Footnote 145
Where states do provide for a remedy against (in)admissibility decisions, it is indeed almost exclusively courts that decide on appeal. However, in the case of more than half of the existing bottom-up instruments of direct democracy, including for example the Swiss proactive initiative and the Dutch agenda initiative, decisions declaring an initiative inadmissible cannot be challenged at all, meaning that the standards of the ICCPR and the Venice Commission Code are not met. Lack of a legal remedy is particularly problematic where a political body, that is, parliament or government, decides on admissibility. Both the Croatian and Latvian constitutional courts have stressed that, where the impartiality of the body deciding on the admissibility of citizens’ initiatives may be contested, availability of judicial review is all the more important.Footnote 146 Nevertheless, parliamentary decisions on formal admissibility cannot be challenged before a court in 12 out of 20 cases, decisions on substantive admissibility in 14 out of 24 cases. Governmental admissibility decisions cannot be judicially reviewed in half of the cases. This means that, in all these cases, no judicial body is involved in the admissibility review at any stage of the procedure. From a rule of law perspective, this is highly problematic.
Timing
In most US states where bottom-up instruments of direct democracy exist, compliance of initiatives with formal requirements is checked before the popular vote, whereas compliance with substantive requirements is not a question of admissibility but may only be challenged after the referendum.Footnote 147 In contrast, all European states review the formal and substantive admissibility of citizens’ initiatives that may lead to a referendum before it may be held. This is in fact what the Venice Commission Code recommends.Footnote 148 There is, however, variance in terms of whether the admissibility procedure takes place before or after signature collection. Furthermore, for some instruments formal admissibility is checked before, and substantive admissibility after the signature collection.
Formal check
In the case of 23 out of 39 instruments that are subject to such a check, the formal admissibility of citizens’ initiatives is reviewed before signature collection (Figure 11).

Figure 11. Timing: formal check
That many states provide for an early check of formal admissibility is understandable. After all, the point of formal checks is to ensure that direct-democratic requests are within the grasp of voters: that they can freely form an opinion about them and express their will accordingly. This freedom may not only be at stake once it comes to a popular vote but already at the signature collection stage. The wording of a given initiative may considerably influence voters’ decisions as to whether to sign it. It therefore makes sense that unclear or even misleading initiatives are filtered out at this early stage.
Substantive check
The picture for the substantive check is just the opposite to that for the formal one: in 23 out of 40 cases substantive admissibility is reviewed after signature collection (Figure 12). Since the point of substantive checks is to remove certain issues from the reach of voters, rather than to ensure the freedom to vote, an early review is generally regarded as less important. That a given citizens’ initiative touches upon a prohibited subject matter does not, as such, affect potential signatories’ freedom to form an opinion about it. Voters may well want to manifest their support of a proposal regardless of whether it might later turn out to be inadmissible on substantive grounds. Reasons of efficiency and cost saving also speak for checking the admissibility of initiatives only once the required number of signatures have been collected. For example, the Swiss Federal Parliament rejected a proposal to review the admissibility of proactive initiatives before instead of, as is currently the case, after the signature collection. A majority of the members of parliament thought that it would not make sense to perform an extensive and costly review of initiatives that then turn out not to gather the necessary support.Footnote 149

Figure 12. Timing: substantive check
Nevertheless, for 17 instruments substantive admissibility is reviewed before the signature collection may start. This is the case, for example, in Slovenia, where the Constitutional Court has stated that it was not least in the interest of voters that any doubts regarding the admissibility of a rejective initiative should be resolved as soon as possible.Footnote 150 Similarly, Liechtenstein has brought forward the substantive admissibility check for the proactive initiative from after to before signature collection as it was thought that late inadmissibility decisions could upset initiators’ expectations.Footnote 151
Assessment
Given the importance of formal admissibility requirements in ensuring the freedom to vote, it is highly problematic that for 16 instruments compliance with them is only checked after the signature collection has been completed. The freedom to vote protects not only voting as such but also acts such as signing citizens’ initiatives. The Italian Constitutional Court stressed as early as 1978 that it must be ensured that voters have a clear understanding of what exactly an initiative they consider signing would entail. The court, therefore, reprimanded the legislator for not having established a system for reviewing rejective initiatives for their compliance with formal requirements before the collection of signatures may start.Footnote 152
If – as is the case in all European states – compliance of citizens’ initiatives with substantive requirements is reviewed as a question of admissibility, then that review should, for rule of law reasons, also be carried out prior to the signature collection. As explained above, especially parliament or the government may be influenced by the popularity of a given initiative when deciding on its admissibility. Once a proposal has been signed by a significant share of voters, it may become difficult for those vying for electoral support to invalidate it. The impartiality of the decision-making authority, a key element of the rule of law, may be strengthened by providing that the admissibility check takes place at a time when there will normally be less political pressure. Of course, reviewing initiatives before signature collection may entail an increased workload for the institution in charge of the review. Yet if this is what the rule of law requires, then that price is to be paid.
Conclusion
As Norberto Bobbio pointed out, democracy depends on a set of basic rules that regulate collective decision-making: for decisions taken by several individuals together to be accepted as collective decisions, they must be taken in procedures that follow clear legal rules which have been established in advance.Footnote 153 Whether and how a citizens’ initiative may be launched, or when and on what topic a popular vote is to be held, cannot be left to the discretion of the authorities. True direct-democratic participation is only possible within a regulatory framework that is based on the rule of law. Admissibility requirements for citizens’ initiatives must form part of such a framework – but only to the extent that they do not unreasonably restrict, but rather enable, direct-democratic participation.
Admissibility requirements have been imposed on 55 of the 57 bottom-up instruments of direct democracy that exist across Europe. Citizens’ initiatives that fail to meet these requirements are filtered out from the very beginning: they are prevented from being debated in parliament, cannot be submitted to a popular vote, and in some cases are not even allowed to proceed to the collection of signatures. As shown in this article, two basic types of admissibility requirements, serving very different purposes, need to be distinguished.
Formal requirements are meant to ensure the freedom to vote. They follow from the very idea of direct democracy: voters will only be able to genuinely express their will regarding a citizens’ initiative if it is clear, homogenous, and so on. Formal admissibility requirements serve to put direct-democratic requests within the grasp of the people by ensuring that the process of will formation lives up to central demands of the rule of law, including foreseeability and transparency. Seen from this perspective, it is highly problematic that for eight of the bottom-up instruments existing across Europe (seven of which are agenda initiatives) there are no formal admissibility requirements at all. Whether the formal requirements that do exist in fact serve to ensure true direct-democratic participation, depends heavily on how they are applied and, therefore, on how the respective checks are designed. Where the decision-maker gives more weight to the interests of the authorities than that of voters, formal requirements can turn into serious obstacles for citizens’ initiatives.
Substantive admissibility requirements put certain issues beyond the reach of the people. Some of these requirements, such as those concerning finances and national security as they exist in numerous European states, reflect a paternalistic attitude and, as such, can hardly be classified as ‘reasonable restrictions’ of the right to participate in public affairs guaranteed by Article 25(a) ICCPR. Others, in contrast, seem to be prompted by a concern to protect democracy by upholding the rule of law. Yet defining the types of norms that should be off-limits to direct-democratic decision-making has proven difficult. Requirements of compliance with ‘international law’, ‘fundamental rights’, or ‘fundamental values of a democratic state governed by the rule of law’ leave the decision-making authorities extremely wide discretion.
Given that admissibility requirements play a fundamental role in ensuring proper direct-democratic participation but, at the same time, are difficult to define and interpret, it is all the more important that states establish procedures for reviewing compliance of initiatives with these requirements that meet rule-of-law standards. Yet for many instruments compliance is not systematically checked at all. Where checks do exist, they are often entrusted to institutions, such as parliaments, that lack legal expertise and/or whose impartiality may be doubtful. The fact that, for many instruments, the admissibility of initiatives is only reviewed once signatures have been collected, heightens concerns that the impartiality of decision-makers might be impaired. Moreover, initiators and voters are only rarely granted the right to participate in the relevant procedures or to appeal the resulting (in)admissibility decisions. Where the design of admissibility procedures does not live up to the demands of the rule of law, there is an increased risk that the existing admissibility requirements will be applied in an unreasonably restrictive – or, on the contrary, a too lenient – way.
Although relevant figures are difficult to come by, initial research within the LIDD project suggests that the share of citizens’ initiatives that are found not to meet formal or substantive admissibility requirements varies greatly, depending on the state and the instrument. In Switzerland, for example, only 1% of proactive initiatives have been declared inadmissible over the last 30 years, whereas in Hungary the respective figure is 93%.Footnote 154 It stands to reason that inadmissibility rates are heavily dependent on how admissibility requirements and procedures are designed. Whereas in Switzerland the only substantive requirement is compliance with peremptory norms of international law, which is reviewed by parliament only once signatures have been collected, Hungary has a very long list of substantive requirements, and it is the National Election Commission that decides on admissibility prior to signature collection. It needs to be left to further research to confirm the exact extent to which the various parameters analysed in this article influence inadmissibility rates. Other factors, such as a country’s democratic and legal culture, might play an equally important role.
Acknowledgments
I am grateful to all the members of the LIDD team – Anna Forgács, Stephania Karasamani, Mara Labud, Fernando Mendez, Caroline Rausch, and Nils Reimann – for their support with the collection and analysis of the data for this article. Thanks are also due to the participants of the workshop on ‘Elections and the International Rule of Law’, held by the ESIL Interest Group on European and International Rule of Law in August 2023 in Aix-en-Provence, to Raffael Fasel, and to two anonymous reviewers as well as one of the editors of EuConst for their helpful comments.
Financial support
This paper grew out of the research project ‘Popular Sovereignty vs. the Rule of Law? Defining the Limits of Direct Democracy’ (LIDD), which was funded by the European Research Council (ERC) (grant agreement 772160).
Appendix
Table 1. Proactive citizens’ initiatives

Table 2. Rejective citizens’ initiatives

Table 3. Agenda initiatives
