The constitutions of the world are incredibly diverse, and the rules governing what is needed for an amendment vary greatly across countries. In this chapter, I describe a series of constitutional amendment stories with the goal of impressing upon the reader this diversity. I will be describing the institutions and providing information about particular events that bring them into relief. Sometimes these events will refer to constitutionally permitted initiatives, while other times they will be stretching the existing constitutional rules (therefore pushing the countries further away from a democratic status). Also, some amendment rules in US states are quite original and, as such, deserve to be included in a chapter on the diversity of amendment rules even though we will not deal with US states in other chapters. I will divide the chapter into three sections.

In Section 1.1, I present the elements and combinations of amendment rules (simple procedures, multiple bodies, qualified majorities, referendums, time, or other constraints) – or, in other words, the additive or alternative combinations that create more complicated pathways for amendments.

In Section 1.2, I argue that these procedures are so convoluted, or that interests are so confrontational, that they can generate institutional conflicts among different constitutional players regarding what the appropriate process is.

In Section 1.3, I will show that sometimes these procedures may create constitutional revisions with the immediate goal of amending the amendment rules themselves so that ordinary amendments will become either feasible or impossible.

This division is not to be understood as creating mutually exclusive and collectively exhaustive categories because each one of the stories I will tell could be classified in different positions. For example, when we have alternative institutional paths for amendment (a case belonging in Section 1.1), it is possible that the actors involved will disagree on which route should be selected (turning it into a case for Section 1.2), and it is also possible that down the road the institutions are changed so that these conflicts will be eliminated (becoming a case for Section 1.3). Depending on what I want to focus on, I will classify the case in one of the three categories, and the reader should know that none of my stories has an exclusive belonging.

These complications, once identified and enumerated, will be compared in Chapter 2 where we will study the constitutional amendment rules of all democratic countries and assess constitutional rigidity.

1.1 Elements and Combinations of Amendment Rules

1.2 Conflicts among Constitutional Players

In India, the requirement for amending the constitution differs depending on the significance of the proposed amendment, with a simple majority of both the upper and lower houses required for basic issues and a two-thirds majority for more consequential ones (The Constitution of India, 1947). Since independence in 1947, there have been numerous conflicts between lawmakers and the Indian judiciary. Starting in 1950, the Supreme Court and the Indian government struggled over the issues of land rights and affirmative action for underprivileged classes. On the issue of land reform, the Indian government had passed legislation that would expropriate the zamindars, a class of landlords who had been given the right to tax small farmers in their jurisdictions during the British colonial period. To compensate the zamindars, the government would pay them future profits from the expropriated land rather than an immediate lump sum. The zamindars challenged their expropriation, and several high state courts struck down the legislation on the grounds that the compensation proposed by the government was not sufficient to make up for the violation of the zamindars’ rights (VanderMay Reference VanderMay1996, Neuborne Reference Neuborne2003, Roy and Swamy Reference Roy and Swamy2022).

In response, the government amended the constitution to restrict judicial review on land-reform cases like those concerning the zamindars. Although this amendment was upheld by the Supreme Court, a subsequent challenge of inadequate compensation by the zamindars in 1952 and later in 1962 led to judicial action: in the 1952 case State of Bihar v. Kameshwar Singh, the Supreme Court struck down “the Bihar Land Reform Act, despite the provisions … removing it from judicial scrutiny, holding that a judicially enforceable just-compensation obligation survived the first amendment” (Neuborne Reference Neuborne2003: 487), and in the 1962 case Karimbil Kunhikoman v. State of Kerala, the Supreme Court struck down Kerala’s zamindari expropriation act. The government then again amended the constitution, further expanding the scope of agrarian reform acts that shielded it from judicial review.

On the issue of affirmative action, the government passed legislation in 1951 that would reserve slots in government-funded educational centers for “untouchables and other backward classes.” These efforts were swiftly challenged by members of the upper-class Brahmins on the grounds that they “violated the fundamental right of equality protected by [the constitution]” (Neuborne Reference Neuborne2003: 488). In State of Madras v. Dorairajan, the Supreme Court considered the Brahmin argument and agreed, striking down the government legislation. As in the land-reform cases, the government responded by amending the constitution to explicitly shield affirmative action legislation from judicial review.

A pattern had been established, where “a state would pass land-reform legislation; it would be challenged in a high court as violating a fundamental right; the challenge would be upheld, and the law declared unconstitutional; [and] there would be an amendment to the Constitution to protect the legislation” (Roy and Swamy Reference Roy and Swamy2022: 25). This pattern would persist for nearly two decades until 1967 when the Supreme Court considered Golak Nath v. State of Punjab. In response to the Punjab Land Reform Act, the zamindars argued that the “act violated articles 31, 19, and 14 [of the constitution], and that the provisions of the … amendments … purporting to shield the act from judicial review … were themselves unconstitutional” (Neuborne Reference Neuborne2003: 489). In a landmark decision, the Supreme Court ruled in favor of the zamindars, holding that amendments to the constitutions were technically laws and, as such, were subject to judicial review in the same way that laws were. While they did not strike down the nationalization-shielding amendments, they threatened to use their new powers to strike down constitutional amendments against any new attempts by the government to undermine fundamental rights.

Following this decision, the Supreme Court proceeded to strike down other government acts, including a bank-nationalizing act, on the grounds that they had provided inadequate compensation. Following these strikes, Indira Gandhi campaigned with the promise to weaken the Supreme Court’s power. She won a super-majority, with 350 seats in the lower house, and made a series of constitutional amendments to wrest the power of judicial review from the courts. These amendments were tested in Kesavananda Bharati v. State of Kerala, and “by a majority of ten to three, the court overruled Golak Nath … [holding that] an amendment of the Constitution was constitutional law which is to be distinguished from ordinary law” (Nanda Reference Nanda1974). However, the victory for Gandhi was not complete: the court also ruled that “parliament’s power of constitutional amendment was not unlimited and that, through judicial review, the limits were to be enforced” (Nanda Reference Nanda1974: 868), arguing that certain core features of the constitution such as fundamental rights could only be revised by wholesale constitutional change. In doing so, the Indian Supreme Court effectively gave itself the power to protect the “core” of the constitution.

In Israel, on July 24, 2023, the so-called reasonableness bill passed the Knesset to limit the Israeli Supreme Court’s power in general. The reasonableness bill curbs judicial review over legislation by explicitly legislating against the Supreme Court’s exercise of judicial review of Basic Laws and requiring a full bench of Supreme Court justices to preside over any case in which the legality of regular legislation passed by the Knesset is evaluated. Under this bill, 80 percent of the bench is required to rule for invalidation of such legislation.

The Israeli coalition government that proposed the reform consists of seven parties – Likud, United Torah Judaism, Shas, Religious Zionist Party, Otzma Yehudit, Noam, and National Unity – and is led by Benjamin Netanyahu, who is the chairman of Likud. According to McKernan (Reference McKernan2023), “the changes are spearheaded not by the prime minister, but by his Likud party colleague Yariv Levin, the justice minister, and the Religious Zionist party lawmaker Simcha Rothman, who chairs the Knesset’s law and justice committee.” McKernan continues to reason that “Levin and Rothman have a longstanding hatred of Israel’s Supreme Court, which they see as too powerful and as biased against the settler movement, Israel’s ultra-religious community, and the Mizrahi population, Jewish people of Middle Eastern origin. In particular, many on the Israeli right have never forgiven the Court for decisions related to Israel’s unilateral withdrawal from the Gaza Strip in 2005.” In 2005, the Israeli Supreme Court had ruled the plan by then–prime minister Ariel Sharon to be constitutional, pathing the way for 9,000 settlers to be evacuated from Gaza. In a ten-to-one vote, it rejected the settlers’ arguments that their human rights would be violated and held that it was ultimately a political decision by stating “their removal had been mandated legally by Parliament and ‘appropriate compensation’ had been ensured by law” (New York Times 2005). Leaders of the right wing contended for decades that the Supreme Court is biased against the settlement movement and acted particularly contradictory when it struck down legislation as unconstitutional, such as in 2020 when the law would have allowed for the retroactive legalization of around 4,000 Jewish homes built on occupied West Bank land privately owned by Palestinians (Halbfinger and Rasgon Reference Halbfinger and Rasgon2020). Therefore, the reform package had initially included an “override clause,” which would have allowed the Israeli Knesset to reenact laws with a simple majority that the Supreme Court had nullified, reducing the court’s ability to strike down the Knesset’s laws that were deemed unconstitutional. Furthermore, it involved granting the government control over judicial appointments by changing the makeup of the Judicial Selection Committee and limiting the authority of its legal advisors by giving the ministers power to appoint and dismiss them, making them subordinate directly to the ministers rather than to the Justice Ministry’s professional oversight. To the government, “these moves are a legitimate way to address a longstanding power imbalance between an overactive and unelected judiciary that selects its own members and that holds unreasonable veto power over democratically chosen representatives” (Kingsley Reference Kingsley2023b). Prime Minister Netanyahu, although not the initiator, backed the reform for supposedly two reasons. First, as he is on trial for alleged corruption, he might personally benefit from the reform “in terms of the administration of his trial” (Goldenberg Reference Goldenberg2023). He denies the charges and has “campaigned against the justice system” since the indictment in November 2019 (Debre and Federman Reference Debre and Federman2023). The second strategic consideration is that he needed the support of the far-right parties to secure the survival of the coalition and, hence, his own power. Mansoor (Reference Mansoor2023) specifically mentions the national security minister, Itamar Ben-Gvir, leader of Otzma Yehudit, who “gave Netanyahu what proved to be the decisive margin.” The article, published in March 2023, points out that Netanyahu agreed to delay the proposal following mass protests and “can’t back away from the proposed judicial reforms entirely without risking Ben-Gvir’s critical political support.”

To the protestors, the reform “will undermine Israel’s democracy by giving absolute power to the ruling coalition and leave minorities without protection from the will of the majority” (Lieber and Boxerman Reference Lieber and Boxerman2022). The protests in March paralyzed the nation, compelling the government to discard its initial strategy of pushing through all parts of the overhaul at once, as already mentioned (see also McKernan Reference McKernan2023). Instead, they are now being introduced through a series of smaller bills, with the override clause having been removed as a sign of conceding to the protesters (Lieber and Amon Reference Lieber and Amon2023). The reasonableness bill still triggered so much controversy – not only on the streets but among the political parties in the Knesset – that all members of the opposition left the chamber when the reasonableness bill was voted. It then passed with a sixty-four to zero vote after all members of the governing coalition voted for it. While the country was waiting for the Supreme Court to decide whether to strike down the reasonableness bill and thereby decide its own fate (Kingsley Reference Kingsley2023a), the reform was shelved like all non-security legislation in October 2023 when Netanyahu was pressured to form an emergency unity government with his main political rival Benny Gantz, leader of the center-right National Unity party, due to the outbreak of the Israel–Hamas war (Hendrix Reference Hendrix2023). The Supreme Court struck down the provision as unconstitutional on January 1, 2024, most likely ending the controversy because the current government cannot address the issue, and it is unlikely that a future government will be able to reintroduce it.

In Hungary, where two-thirds support from the National Assembly is required for an amendment to be approved, a number of constitutional changes followed after the election of the Fidesz party in 2010 to the Hungarian Parliament with a supermajority. Fidesz came to power in the context of the economic fallout that had followed the 2008 worldwide recession. Voters rejected the Hungarian Socialist Party, which had been in power since 2002. Although the Fidesz party won only 53 percent of the vote, this translated into a supermajority of seats in Hungary’s unicameral parliament (Bánkuti et al. Reference Bánkuti, Halmai and Scheppele2015). This disproportionate result was due to the Hungarian electoral system, which assigned 386 legislative seats using a combination of single-member districts, proportional representation-style lists, and fifty-eight compensatory seats originally designed to make up for the disproportional allotment of single-member districts (Benoit Reference Benoit2001). Fidesz’s supermajority allocation of seats despite only winning a bare majority of votes was a result of its high degree of support in smaller, rural regions, where votes are translated to seats more efficiently, and due to the fact that because only three parties ran in the 2010 election it benefited from a relatively large proportion of the compensatory seats (Bánkuti et al. Reference Bánkuti, Halmai and Scheppele2015).

After laying the groundwork during their first year in government,Footnote ¹⁰ Fidesz introduced a new constitution called the Fundamental Law in 2011 and proceeded to amend it significantly several times over the next few years. In June 2012, the first amendment to the Fundamental Law elevated transitional provisions to bona fide parts of the constitution in order to shield them from judicial review, ensured that the presidential salary could only be changed via the adoption of a Cardinal Law (a law whose passage requires a two-thirds supermajority), and removed a clause in the transitional provisions weakening financial oversight in response to EU criticism. The transitional provisions were a set of amendments that had not been technically ratified between the passage of the Fundamental Law in 2011 and the first round of amendments in 2012. They included clauses that guaranteed there would be no statute of limitations on crimes committed during Hungary’s communist regime, that removed the chairs of the Supreme Court, National Judicial Council, and privacy ombudsman, and that guaranteed that fines levied against Hungary by the European Commission can be collected as taxes (Boros Reference Boros2013, Roznai Reference Roznai and Law2022).

In October 2012, the second amendment modified the transitional provisions to make voter registration in a government-controlled database prior to an election a mandatory requirement for voting. In December 2013, in response to a claim filed to the Constitutional Court that the transitional provisions were not temporary and were in fact unconstitutional attempts to amend the constitution, the Constitutional Court struck down this amendment as well as a set of other clauses from the transitional provisions that Fidesz had tried to elevate to Fundamental Law, such as the transitional provision that removed the statute of limitations on crimes committed during communism. This decision was made on the grounds that the transitional provisions were indeed permanent and that “their adoption exceeded the government’s power under the Fundamental Law to enact transitional provisions, notwithstanding the government’s clear intent to give them constitutional status” (Roznai Reference Roznai and Law2022: 150).

The Fidesz government responded with the Fourth Amendment in March 2013, which sought to undermine the Constitutional Court’s decision by directly incorporating the transitional provisions into the Fundamental Law itself. Critically, it removed the ability for the Constitutional Court to review constitutional amendments on substantive (rather than strictly procedural) grounds and prohibited it from referring to precedents in constitutional law from before the 2011 constitution. In addition, the amendment stipulated that political advertising during campaigns could only be broadcast over public media sources with the list of acceptable media groups to be determined by Cardinal Law, that speech which violated the “dignity of the Hungarian nation” could be criminally prosecuted, that those who received tuition grants from the government would be able to work at Hungarian companies thereafter, and that municipal governments could use the powers at their disposal to remove homeless populations from public spaces (Orange Files 2013). The effort to shield the constitution from judicial review was effective: later in the year, the Constitutional Court “rejected a challenge to various provisions of the fourth amendment … [on the grounds] that it lack[ed] competence to conduct substantive judicial review of constitutional amendments due to explicit provisions, introduced by the Fourth Amendment, that limit[ed] its jurisdiction over constitutional amendments to formal review on procedural grounds” (Roznai Reference Roznai and Law2022: 152).

In Slovakia, where a three-fifths qualified majority of legislators is needed to change the constitution, a 2014 constitutional amendment to impose restrictions on the qualifications of new and incumbent judges passed but was challenged by the Constitutional Court. If passed uncontested, the amendment would have required all new and existing judges to meet a certain degree of security clearance in order to be appointed or to retain their positions. On January 30, 2019, the court struck down the 2014 amendment on the grounds that it violated Article 1 of the Slovak Constitution, which held that the Slovak Republic was a democratic state bound by the rule of law. The court argued that it directly followed that the constitution ensured commitment to the separation of powers and the independence of the judiciary, which it considered to be “material core of the Constitution” (Domin Reference Domin2019). By this ruling, the Slovakian Constitutional Court therefore gave itself the power to reject constitutional amendments that go against the “core of the constitution” (Ľalík Reference Ľalík2020a).

The judges made the decision shortly before nine out of thirteen would leave, and it almost went unnoticed “because the political fight over the composition of the future Constitutional Court took center stage” (Ľalík Reference Ľalík2020a: 328; see also Steuer and Láštic Reference Steuer, Láštic and Pócza2024: 248), who describe the period from 2018 to 2019 as a “gridlock in electing new judges.” A few months later, the “Threema” scandal surfaced after the murder of the investigative journalist Ján Kuciak and his fiancée. Nineteen judges were accused of “delivering judgments on demand” (Čuroš Reference Čuroš2023: 639), painting a picture of “a massive corruption scheme among the judiciary, law enforcement, politicians, and business members” (Čuroš Reference Čuroš2022). As a result, “the outrageous crisis of legitimacy of the judiciary provided a perfect opportunity for the interference of the legislative and executive power into the judicial power. Such interventions had been attempted for years, but after the Threema scandal, the public embraced the idea of enacting significant changes in the judiciary” (Čuroš Reference Čuroš2022). In this light, some view the reform that followed favorable, leading to more “efficiency and a better standard of delivering decisions” (Čuroš Reference Čuroš2022), while others point toward the illegitimacy: “The passing of the amendment in question has made the parliament an unbound constitutional-maker that is neither in line with constitutionalism nor with democracy” (Ľalík Reference Ľalík2020b).

The reform, curtailing the power of the judiciary, was passed in December 2020. Steuer and Láštic (Reference Steuer, Láštic and Pócza2024) state that “the amendment can be seen as a parliamentary retaliation to the SCC’s [Slovak Constitutional Court’s] decision” (255). The amendment allowed for the examination of the origins of the property holdings of individual judges (purportedly to investigate corrupt dealings by judges), eased the process of judicial appointment to the Constitutional Court by allowing a simple majority of the Slovakian Parliament (rather than a three-fifths supermajority) to appoint a constitutional justice, provided the president the power to unilaterally appoint justices if the parliament fails to do so, and mandated a maximum retirement age of seventy-two years for judges on the Constitutional Court (Davala and Chudo Reference Davala and Chudo2021). In addition, and most critically, the amendment removed the ability of the court to review constitutional amendments and was rushed through Parliament, just barely passing the three-fifths threshold with 91 of 150 MPs voting in favor (Steuer and Láštic Reference Steuer, Láštic and Pócza2024).

In 2021, the Slovakian Supreme Court made a significant ruling under the new rules, stating that a referendum initiative on a snap election was unconstitutional. “The Court found referenda analogous, both in form and function, to constitutional amendments” (Drugda Reference Drugda2021), thereby reclaiming some of the power it had lost through the amendment in 2020.

This absence of a mechanism for early elections became acutely salient in 2022 when President Zuzana Caputova responded to the collection of over 380,000 signatures by three opposition parties and decided to hold a referendum, seeking approval for the early termination of the National Council of the Slovak Republic’s election period either through a referendum or a resolution by the National Council. Shortly after that decision, Prime Minister Heger’s coalition government faced a vote of no confidence and subsequently collapsed (Associated Press Reference Smyth2023a).

The referendum was held on January 21, 2023, and faced challenges due to the president stating, “I perceive this referendum as a part of a political campaign of one party … I am therefore not going to encourage citizens to attend, nor am I going to discourage them” (Cincurova et al. Reference Cincurova, Inotai, Gosling and Ciobanu2023). Subsequently, it resulted in a low turnout and its failure (as at least 50 percent would have been needed).

In the week prior to the referendum, President Caputova had already given parliament the deadline of the end of January to amend the constitution to make a snap election possible. Only four days after the failed referendum, ninety-two members of parliament voted for the constitutional amendment that allows a snap vote if it is approved by a three-fifths majority in the 150-seat National Council of the Slovak Republic (Associated Press 2023b).

In Ukraine, the Constitutional Court struck down constitutional amendments as unconstitutional years after they were made. In 2004, after the Orange Revolution, a major constitutional change was made to limit the power of the president relative to the prime minister. This amendment was proposed and passed by parliamentarians supporting the pro-Russian Viktor Yanukovych, ostensibly with the goal of weakening the power of the anti-Russian president Yushchenko. In 2010, Yanukovych regained the presidency and appointed four new members to Ukraine’s Constitutional Court. Subsequently, the court ruled that the 2004 amendments to weaken the presidency were unconstitutional and ordered that they be reversed, restoring the power of the president. Following the widespread Euromaidan protests in 2013, Yanukovych fled the country, new elections were held, and the newly elected parliament repassed the 2004 amendments to weaken the presidency (Tyushka Reference Tyushka2014).

In all of these cases, there are conflicts between the legislature and the judiciary, and in lots of them, there is an oscillation in the results: sometimes the legislature is the winner, and in others it is the courts (India and Israel are the most prominent examples). What is not obvious is the (indirect) participation of the public because, given that the constitution does not specify who should be the ultimate decisionmaker, it is the actor that will express the opinion of the public that will claim the constituent power. The following case demonstrates what happens when the carrier of the constituent power can interfere directly. It is a conflict not between legislation and constitutional interpretation (conflict within the constitutional rules) but between interpretation and constitutional amendment.

In a landmark case in October 2006, Ireland’s high court rejected the recognition of a lesbian couple’s Canadian marriage. It ruled that the constitution did not protect the rights of same-sex couples to get married, although the constitution had never defined marriage as being between a man and a woman explicitly. Instead, “[Justice Elizabeth] Dunne took the definition of marriage contained in Ireland’s Civil Registration Act 2004 which defines marriage as being between a man and a woman as an indication of the ‘prevailing view’ as to the definition of marriage” (Tiernan Reference Tiernan2020: 31). The plaintiffs, Katherine Zappone and Ann Louise Gilligan, had challenged an interpretation of the Taxes Consolidation Act, which provides for a husband to be assessed on his and his wife’s total income and vice versa, by the Revenue Commissioners, who had denied altering their tax returns in recognition of their marriage explicitly. They had based the interpretation on the definitions of husband and wife in the Oxford English Dictionary, and Zappone and Gilligan argued to the court that it violated their constitutional rights to equal protection (Article 40.1), privacy, and dignity (Article 40.3). Justice Dunne recognized that “it is to be hoped that the legislative changes to ameliorate these difficulties will not be long in coming” (High Court 2006). Also, she delegated the responsibility for such change to happen by saying that “ultimately, it is for the legislature to determine the extent to which such changes should be made” (High Court 2006).

In response to the ruling, lower house member Brendan Howlin tabled a Private Member’s Civil Union Bill on the same day that Zappone and Gilligan lost their high court case. Even though the bill was defeated in Parliament later, the introduction of (various) civic partnership bills in conjunction with this court case raised public awareness about the need for legal recognition for same-sex couples (Tiernan Reference Tiernan2020: 39). When the Civil Partnership and Certain Rights and Obligations of Cohabitants Act was finally passed in 2010, it was clear that a constitutional amendment was needed to achieve full equality in terms of marriage and not only in terms of partnership. In 2012, a constitutional convention was established due to complex government formation negotiations between the center-right Fine Gael and the center-left Labour Party over their differing reform agendas regarding the issue of same-sex marriage, among other issues. The convention’s agenda “reflected the decision of the inter-party negotiators to ‘park’ certain matters that were in their respective election manifestos that were unlikely to be resolved easily during their febrile and intense negotiations” (Farrell et al. Reference Farrell, Harris, Suiter, Contiades and Fotiadou2016: 122). After the convention recommended amending the constitution, the government called a referendum.

The referendum on May 22, 2015, had a remarkably high turnout. A majority of 62 percent voted to add the following sentence to the Irish Constitution: “Marriage may be contracted in accordance with law by two persons without distinction as to their sex,” making Ireland the first country in world history to extend civil marriage to same-sex couples through a popular vote. Eventually, the people overruled the court and its interpretation of the constitution. However, Doyle and Walsh (Reference Doyle and Walsh2020) emphasize that the outcome of the referendum was not particularly contentious or surprising:

In addition to this case, there were eight others in Ireland where a proposed amendment was “regarded as necessary in order to reverse statements of law resulting from unpopular judicial interpretations, or because of a judicial decision making it clear that a desired course of action would be possible only following a successful referendum to amend the Constitution” (de Londras and Morgan Reference De Londras, Morgan and Contiades2013: 182). Among the eight proposed amendments, one was on human rights (bail), two on elections, one on abortion, and four on institutions of government. Concerning bail, Raifeartaigh (Reference Raifeartaigh1997) states that “Ireland appears to be one of the few, perhaps the only, common law jurisdiction which rejects the likelihood of further offending as a ground for pre-trial detention. In 1966, the Irish Supreme Court emphatically rejected the suggestion that one of the grounds on which bail might be refused was the likelihood that the accused might commit further offenses; again in 1989 it unhesitatingly affirmed its earlier decision. The Court’s view, as we shall see, was founded on the view that such a course of action would violate the presumption of innocence” (2). What Raifeartaigh calls “a somewhat absolutist view of the presumption of innocence” led the government to hold a referendum to change the Irish legal position (Raifeartaigh Reference Raifeartaigh1997: 18). The new article, approved by the people in 1996, ensures that the grounds for refusal of bail by a court may be regulated in the future by the legislature so that, under the Bail Act 1997, a court can now consider whether or not a person had committed serious crimes while on bail in the past in addition to the “extraordinary circumstances” that the ruling in 1966 had implied (Raifeartaigh Reference Raifeartaigh1997: 18).

Two out of these eight cases failed in the respective referendum: the one on representation of rural voters and the one on parliamentary inquiries.

1.3 How Institutions Are Modified in Order to Get the Intended Outcome

In this section, I will highlight some examples from US state constitutions because they have provisions that involve referendums which, as will be explained in the next chapter, are very flexible instruments for changing the rules. The examples in this section demonstrate that modifying how easy it is to use referendums affects the facility of constitutional amendments.

The Michigan Constitution from 1850 included the provision that the voters must decide whether a constitutional convention is called automatically every sixteen years. Furthermore, the provision required that a majority of votes that are cast in the election – not just a majority of those voting on the question of the convention call – must approve all convention calls rather than only those that occur automatically. Only in 1866, pursuant to the sixteen-year requirement, and in 1906, when it was placed on the ballot by legislative action, did the voters approve the calling of a constitutional convention. In many instances, the question failed despite the majority of those voting on the question giving their approval (“CRC special report” 2010a). Five referendums failed between 1926 and 1961, and, especially during and after World War II, “a general dissatisfaction with the document created a growing desire to revise the constitution” (“CRC special report” 2010a: 2). Therefore, instead of a general constitutional revision, eighteen proposed amendments were adopted in the decade beginning in 1951, while only three were rejected (“CRC special report” Reference Carrubba, Gabel, Helmke, Martin and Staton2010b). The Citizens Research Council of Michigan summarized the developments in this era as follows: “Along with 51 previous amendments, these new provisions gave the Constitution the appearance of a patchwork quilt of trivia and excessive detail, which provided for far too many executive branch agencies, excessive earmarking of taxes, and a system of legislative representation skewed toward rural interests” (“CRC special report” Reference Carrubba, Gabel, Helmke, Martin and Staton2010b: 4).

As a consequence, the “long struggle for constitutional revision” (Cramton Reference Cramton1964: 8), amplified by the fiscal crisis in 1959, stimulated renewed efforts by a number of interest groups that had previously supported the campaigns (Sturm Reference Sturm1963). In January 1960, the Michigan Junior Chamber of Commerce and the Michigan League of Women Voters proposed a constitutional amendment to ease the calling of a constitutional convention. The so-called Gateway Amendment was approved by the voters later in 1960 and paved the way for a successful convention call in 1961. The referendum on April 3, 1961, did not require the majority of electors anymore, and therefore a favorable plurality on the convention question was sufficient. Michigan voters approved the proposition by a margin of only 23,421 votes (Sturm Reference Sturm1963). The Citizens Research Council of Michigan emphasized that “if the former constitutional requirement of a majority of those participating in the election had applied, the proposal would have failed” (“CRC special report” 2010a).

To summarize, changing the rules in 1960 was critical in allowing a convention referendum to pass in 1961 after it had been blocked on various occasions and only stepwise amendments had been possible.

In Ohio and other states, lawmakers were recently trying to achieve the opposite: They were pushing for a rule change that would make it harder for citizens to initiate and implement constitutional amendments successfully. The rule change, sought by the Republican supermajority, was specifically targeted at a citizen-led effort to put a constitutional amendment on the ballot that would prohibit banning abortion before fetal viability (Zernike and Wines Reference Zernike and Wines2023). The amendment that the lawmakers advanced had already passed the senate and the house. It went before voters in a special election in August 2023 and thus would have taken effect before the amendment on abortion rights advanced by citizens was on the ballot in November 2023.

First, the addition of new requirements to get proposed amendments on the ballot would have immediately hampered the citizens’ effort: according to the New York Times, “proponents would have to collect signatures from at least 5 percent of the residents in all 88 counties in the state, up from the current 44” (Zernike and Wines Reference Zernike and Wines2023: para. 29). Additionally, they would not have been allowed to collect additional signatures to make up for those that authorities disqualified, which as of now can be done for one week (the so-called curing period) (Zernike and Wines Reference Zernike and Wines2023). Second, their ballot initiative would have required a 60 percent threshold once it passed the higher hurdles to appear on the ballot, rather than the current 50 percent. The Republican measure to raise this threshold, however, required support from only 50 percent of voters to pass. Furthermore, Chris Melody Fields Figueredo, executive director of the Ballot Initiative Strategy Center, which works to support progressive ballot measures, expresses suspicion in the New York Times that the new threshold was chosen strategically based on the vote for abortion rights in other red and purple states (Zernike and Wines Reference Zernike and Wines2023). Specifically, in Michigan, Kentucky, and Kansas the vote was between 52 and 59 percent.

However, 57 percent of the Ohio voters resoundingly rejected the amendment of the amendment rule (Ingles and Kasler Reference Ingles and Kasler2023). Subsequently, abortion rights were enshrined into the state’s constitution in the November referendum by 57 percent (so not the 60 percent that would have been necessary under the new rules), although the approval has prompted another institutional conflict over the amendment’s implementation. How and when the impacts of the new constitutional protections for abortion access and other reproductive rights are felt therefore “remains unclear” (Smyth Reference Smyth2023b).

Similar measures have been taken in North Dakota in response to issues other than the abortion rights movement: “North Dakota lawmakers in recent years have grumbled about certain constitutional initiatives voters have approved, including measures for a state Ethics Commission in 2018 and for term limits on the governor and state lawmakers last year” (Dura Reference Dura2023). Consequentially, the North Dakota legislature approved a bill in 2023 that makes it more challenging for constitutional initiatives to qualify for the ballot and be adopted. Voters will be asked on the 2024 ballot to approve or reject the constitutional amendment, which would (1) establish a single-subject rule for initiatives (both statutory and constitutional), (2) increase the signature requirement for constitutional amendment initiatives from 4 percent of the resident population to 5 percent of the resident population of the state, and (3) require proposed constitutional initiatives that have qualified for the ballot to win approval in both the next primary and the next general elections (Mitchell Reference Mitchell2023).

In Arkansas, Republican lawmakers chose another route rather than amending the constitution: After voters had soundly rejected a constitutional amendment in 2022, which was proposed by the legislature to stiffen the requirements to get a measure on the ballot (increasing the required vote to approve ballot initiatives to 60 percent and applying to measures placed on the ballot via petition or by the legislature), the legislature simply passed new requirements as state law. They also aimed to increase the hurdles but targeted hurdles other than the ones that had just been rejected, picking up on a rejected amendment from 2020. Governor Huckabee Sanders signed the law in March 2023, whereby the number of counties where a minimum number of signatures from registered voters must be submitted was raised from fifteen to fifty (Smyth Reference Smyth2023a). The minimum is defined as “not less than one-half of the designated percentage of the electors” (Vrbin Reference Vrbin2023), with the percentage being set at ten for proposed constitutional amendments (Smyth Reference Smyth2023a).

Conclusions

The goal of this chapter is to impress upon the reader two points: first, the diversity of constitutional amendments, and second, the significance of them. Section 1.1 deals with this diversity and shows how many different simple rules are used (different actors are assigned the role of a veto player, different decision-making rules are selected, intermediate elections are required, referendums may be added in the mix, or time or other constraints are imposed; then sometimes these rules are combined, and other times they are presented as alternatives). All such cases are presented in Section 1.1, with the goal to underline the diversity and complications of constitutional amendment rules.

Sections 1.2 and 1.3 focus on the significance of these rules by presenting how these rules become the objects of political conflict or the targets of modification for political reasons, making constitutional change easier or more difficult to achieve. These two sections demonstrate that the policy preferences of different actors become the source of induced institutional preferences: The actors understand that the best way to promote their policy preferences is to modify in a specific way the rules of the game (that is, alter the constitution). The conclusion of these two sections is to persuade the reader that these amendment rules matter a lot – a conclusion that may be trivial for some readers but has been disputed in the literature.