Turkey has undergone a dizzying array of crises over the last five years. Beginning with the repressive crackdown against the Gezi Protests during the summer of 2013, the country has gone from being cited as a model Muslim democracy to taking pride of place on the growing worldwide list of democratic reversals. Pundits now lump Turkey’s president, Recep Tayyip Erdoğan, in with populist authoritarian leaders ranging from Hungary’s Victor Orbán to the Philippines’ Rodrigo Duterte. On some indices Turkey leads the pack, jailing more journalists than any other country, throttling the independence of the judiciary and establishing a near total stranglehold on the media.
The country’s descent into majoritarian authoritarianism accelerated in 2016 with a failed coup attempt followed by the imposition of a state of emergency that lasted two years. With the rule of law all but suspended, Erdoğan’s government purged perceived opponents, waged a war against the country’s Kurdish population and stifled all forms of dissent, marking a fundamental break in the country’s political trajectory.
That break has now been institutionalized by means of a constitutional transformation that is perhaps the most audacious gambit of Erdoğan’s 16-year rule. Under cover of the state of emergency, with nearly all democratic prerequisites for a constitutional referendum suspended and troublesome members of parliament and constitutional court judges detained, the governing Justice and Development Party (AKP) put in place a set of constitutional amendments that effectively repealed the democratic character of the republic. And they accomplished this constitutional coup in the name of saving democracy.
Regular elections may continue under Turkey’s amended constitution, but the concentration of power in the executive gives it de facto control over both the legislative and judicial branches, eroding the separation of powers and checks and balances that are the basic prerequisites for democratic rule. Turkey now provides a blueprint for how a partially consolidated democracy may be dismantled from within using constitutional tools.
Constitutional Amendments Before 2017
There is considerable debate about whether this tragedy was foretold—that is, whether the AKP came into office in 2002 with the goal of establishing constitutional authoritarianism. From the outset, constitutional reform was an element of the party’s political agenda but that was true of almost all parties in the Turkish political spectrum in 2002. The constitution adopted in 1982 was drafted under military rule with deeply illiberal and anti-democratic provisions. Widely seen as an obstacle to democratic consolidation in the country, the 1982 constitution was amended nearly 20 times over three decades (from 1987 to 2017). Most amendments came in the form of packages changing multiple provisions at once and many were adopted by broad coalitions of parties across the political spectrum.
Until 2017, amendments to the constitution largely shared a set of common characteristics: They were designed to liberalize, democratize and civilianize the constitution. More specifically, amendments served three ends: enlarging fundamental rights and freedoms for Turkish citizens; strengthening rule of law guarantees constraining state action; and removing or reducing special privileges set aside for the military. Amendments lifted bans on political parties, eliminated security courts, abolished the death penalty, introduced affirmative action and reduced the military’s role in civilian governance.
Once Turkey began to pursue accession to the European Union in the early 2000s, there was broad parliamentary consensus behind adopting liberalizing and democratizing reforms in keeping with EU standards. The earlier amendments routinely earned praise from constitutional experts of the Council of Europe’s Venice Commission for Democracy through Law. There was also widespread consensus, however, that the amendments did not go far enough and that the consolidation of democracy in Turkey would require a new constitution.
But a new constitution was not in the cards. Turkish constitutional identity is marked by two fundamental cleavages that produced an impasse whenever a constituent assembly or constitutional commission was convened to consider wholesale change. Those cleavages—over the status of religion and an ethnic definition of citizenship—mirror the basic polarization in the society between secular and religiously conservative communities and between ultra-nationalists and those who prefer a civic and inclusive definition of citizenship.
The AKP has deftly maneuvered around these cleavages, sometimes exploiting them and sometimes sidestepping them, to press through incremental reforms that have fundamentally redefined the constitutional identity of the republic without entirely replacing its constitution. Six sets of amendment packages were passed under AKP governments prior to 2017. The first four packages prior to 2006 were largely consistent with EU accession requirements, including abolishing the death penalty. In 2007, the AKP embarked on more ambitious structural transformations, which were presented as dismantling anti-democratic elements of the Turkish constitutional order—notably by limiting the role of tutelary institutions put in place by the military junta in 1982 to cabin the authority of the elected branches of government.
But some observers within and outside the country worried that limiting the tutelary role of the military and judiciary—long seen as bastions of secularism and commitment to the founding ideology of the state (known as “Kemalism” after the founding statesman Mustafa Kemal)—might have a double-edged effect, enabling the AKP to erode the Turkish constitutional commitment to secularism. Another source of concern was the ruling party’s insistence on an increasingly majoritarian definition of democracy. Then Prime Minister Erdoğan was especially fond of invoking his electoral mandate to silence critics.
The worry that the 2007 constitutional amendment—which introduced direct elections for president in place of an appointed and largely symbolic president—might open the door for an emboldened majoritarian presidency proved prescient. Seven years later, these fears were fully realized in Erdoğan’s invocation of a populist, majoritarian mandate to concentrate power in the executive once he became the first directly elected president. But such concerns were given short shrift at the time of the 2007 referendum. Following on the heels of an attempt by the military to block the appointment of the AKP’s preferred presidential candidate, the amendment was popular with an electorate fed up with military intervention and passed by a large majority.
More significant concerns were raised with the AKP’s 2010 constitutional amendment package, which cobbled together liberalizing changes—enhancing some individual rights and further restraining the military—with a restructuring of the judiciary that would alter the composition of the bench. Here again, these concerns were largely sidelined due to long-standing grievances over judicial insistence on a restrictive definition of secularism. The judicial system of appointments and promotions imposed an ideological litmus test to ensure that judges would adhere to a strict interpretation of the state’s founding ideology. The de facto effect of the system was to make clear that religiously observant judicial candidates need not apply.
On their face, the 2010 amendments were unobjectionable, removing this ideological constraint and including a greater swathe of the judiciary in the appointment and promotion process. Moreover, a majority of Turks, and especially the AKP’s own constituents, were happy to countenance changes that allowed a more permissive approach to the country’s jurisprudence on secularism. In practice, the amendments would predictably mean that a greater proportion of judges would have ideological affinity to the ruling party, since these were the very candidates previously excluded. Still, because the amendments were in line with broader democratic practice concerning high judicial councils, they won the support of the Venice Commission.
As a result of the other, more liberal features of the package, the 2010 amendments gained the support of constituencies beyond the AKP’s base, passing with 58 percent of the vote (while the AKP has never garnered more than 49 percent of the vote in a general election). Nonetheless there was good reason to be concerned that the AKP was embracing a form of tactical liberalism to facilitate court packing.
In retrospect, however, it was the AKP that regretted the judicial restructuring of 2010. Having strengthened the autonomy of the High Council of Judges and Prosecutors (HCJP) and limited the role of the executive in appointments, the government found itself blindsided in December 2013 by a criminal probe into cabinet ministers and their relatives on corruption charges being investigated by the very judges and prosecutors that were ushered in following the amendments.
Treating the investigation as an attempted judicial coup, and declaring that the judiciary had been infiltrated by its opponents from the Gülen movement, the government immediately introduced legislative changes to limit the powers of the Council while simultaneously removing or transferring hundreds of judges and prosecutors, replacing them with more reliable pro-government appointees. This legislation was manifestly unconstitutional and was eventually invalidated by the Constitutional Court, but since such rulings do not have retroactive effect the individuals reassigned or removed were not restored to their positions. Further legislative changes followed, not only reversing the gains in judicial autonomy from the 2010 amendment package, but politicizing judicial elections to the point of abrogating the independence of the judiciary altogether.
In many ways, the 2010 constitutional amendments and their aftermath were the harbinger of the constitutional transformation to come. The AKP of 2002, positioned on the center right of the Turkish political spectrum, included well-known and non-Islamist politicians under a reformist and anti-corruption political platform, with real promise for both economic and political liberalization. But over its first decade in office, many identified with political reform were sidelined while the core of the party’s political power coalesced around Erdoğan. Whatever liberalizing potential the party might once have had was extinguished as the party became a vehicle for the ambitions of its leader. By 2013 it was clear that the cumulative effect of the AKP’s incremental constitutional reforms had chipped away at democratic norms, clearing the way for Erdoğan’s consolidation of power.
Whereas the trajectory of Turkish constitutional amendments from 1987 to 2010 could plausibly be described as democratizing, the Turkish case represents a cautionary tale of how easily stalled democratic consolidation can devolve into democratic reversal. The tools of constitutionalism can be particularly insidious in these cases, providing an appearance of legitimacy to changes that entrench authoritarian control while vaunting the putative democratic pedigree of reform by referendum.
Constitutional Transformation by Coup
The circumstances of the attempted coup of June, 2016 and its aftermath have been well-described elsewhere. What is critical to understand is the degree to which the failed coup represented a “gift from God,” to quote Erdoğan’s memorable response, for those seeking a pretext to establish emergency rule and consolidate power. The AKP had made the replacement of the country’s parliamentary system with an executive presidency central to its campaign in the general elections of 2015 but the idea received a cold reception even among the party’s own base. With a state of emergency in place, however, the AKP seized on the opportunity to reshape the electoral landscape, tilting the playing field in its favor to realize its constitutional ambitions. Large-scale constitutional change that had faced significant electoral opposition prior to emergency rule was now made possible by suspending the basic prerequisites for a free and fair democratic vote.
The state of emergency included wide-ranging purges of civil servants—including everyone from public university professors to K–12 teachers to appellate court judges, chief prosecutors and senior ministry officials to agricultural inspectors and low-ranking clerks—and the near total prohibition on freedom of assembly other than rallies by the ruling party. The purges all but paralyzed the country’s legal system with even ordinary court cases unable to proceed due to the mass firing of judges, while basic procedural protections were denied those languishing in pretrial detention or convicted in summary proceedings. The expropriation and closure of nearly 200 broadcast and print media outlets and the jailing of journalists as well as opposition MPs, too, was made easier (though MPs had lost their parliamentary immunity even before the imposition of emergency rule). But the convening of a referendum to pass sweeping changes to the constitution was by far the most consequential of the measures taken under emergency rule: The constitutional changes ensured the durability of Erdoğan’s accumulation of powers beyond any state of emergency.
The dramatic scale and scope of the constitutional transformation achieved by the April 16, 2017 referendum is apparent in the measures adopted. The 18 articles of the amendment package resulted in modifications to nearly 50 constitutional provisions and the repeal of an additional 21. The net effect of these changes was to end the system of parliamentary government that characterized the Turkish republic from its founding, replacing it with an executive presidency system that undermines the separation of powers and imposes almost no checks other than periodic, and increasingly orchestrated, elections to hold presidential action accountable.
The specific measures that make up this sui generis executive presidency (or “Turkish-style presidentialism,” as AKP officials sometimes describe it) are a hodgepodge of provisions borrowed from earlier precedents and other constitutional systems, cherry-picking features to maximize the concentration of power in the presidency while eliminating the checks that protect against authoritarianism. Kim Scheppele has referred to this method as the “Frankenstate problem”—a system of government “composed from various perfectly reasonable pieces” but rendered monstrous from “the horrible way that those pieces interact when stitched together.” The 2017 Turkish constitutional amendments may be one of the starkest examples of such a constitutional gerrymander.
The 2017 referendum, first of all, substantially and dramatically weakens the legislative branch in favor of the executive. The number of deputies was raised from 550 to 600 and the age of eligibility was lowered to 18. The council of ministers (the cabinet) and the prime minister’s office were abolished and their powers transferred to the presidency. Powers that the cabinet never possessed—notably authority over the armed forces—are also now invested in the president. Parliament no longer has a role in appointing the cabinet. Other traditional parliamentary powers such as setting the annual budget or regulating the state audit authority are likewise transferred to the president. In addition, the president is empowered to veto legislation and while parliament retains the ability to override a veto such action now requires a higher threshold vote. The powers of inquiry and interpellation by the parliament are also starkly limited, with no provision for questioning ministers and an extraordinarily high voting threshold to open an investigation into alleged criminal responsibility of executive branch officials. The only other accountability lever available to the parliament is to call for new elections—effectively dissolving itself but also requiring a new presidential election—but this power, too, is now subject to an extraordinarily high vote threshold of a three-fifths majority.
Moreover, the constitution designates domains inherently related to executive power to be governed by executive decree. This legislative authority is not delegated by parliament to the executive but rather reserved for the president as a constitutional matter. As a result, there is no need for an empowering legislative framework for presidential decrees, which might have set some limits. While the amendments enshrine the principle that legislation would prevail over presidential decrees in case of conflict, there are reasons to doubt the effectiveness of this principle in practice. With legislative and presidential elections combined on a five-year cycle, and presidential candidates now allowed to be members of political parties, the likelihood that the president will be from the party that commands a majority in parliament is heightened because the president will effectively be running as the top of the ticket in a combined election. If the president’s party commands a majority in parliament, the likelihood of legislators overriding an executive decree with countervailing legislation is low. In short, the constitutional amendments in the area of legislative power transfer competencies to the executive and subject parliament to extensive new constraints.
The measures passed in the 2017 referendum establishing the executive presidency represents an even more far-reaching assault on the separation of powers. The reserved areas in which the executive now has law-making authorities effectively create a parallel system of administrative laws with little opportunity for judicial review or other checks. The original legislative power accorded to the president under Turkey’s new constitutional system, and the absence of parliamentary checks on that authority, is the clearest example of the excessive concentration of executive power. Moreover, this enhanced power is granted to a partisan officeholder.
The presidency was long deemed neutral, non-partisan and symbolic, but since 2015 Erdoğan has openly disregarded the constitutional injunction that the president must renounce party affiliation. The constitutional amendments remove the requirement of non-partisanship, clearing the way for heads of political parties to run for the presidency, making it more likely that the president will lead a party that wields a majority or plurality in parliament. The partisan president is also free to appoint and dismiss cabinet members at will—they need not be elected officials and are accountable solely to the president. Thus, unlike the council of ministers under the parliamentary system made up of elected MPs accountable to the public, the government now answers exclusively to the president. These officials—vice-presidents and ministers—are not of a fixed number, have no set terms of office, and their portfolios and division of responsibilities is entirely at the discretion of the president. Parliament has no power to approve or veto nominations. The president is also free to select some ministers from parliament, creating a significant patronage opportunity with the legislature.
The powers of the presidency enumerated within the amended provisions of the constitution include: the power to appoint and dismiss vice-presidents, ministers and senior state officials; the power to legislate by presidential decree on executive matters; the power to determine national security policies; the power to appoint the chief of the general staff of the military and the power to appoint, and regulate by presidential decree, the National Security Council; the sole power to declare a state of emergency and the power to issue emergency decrees for its duration; the power to dissolve parliament and call early elections; the power to prepare the state budget; the power to veto laws; the power to address parliament on matters of domestic and foreign policy; and the power to appoint members of the Council of Judges and Prosecutors and to appoint judges to the Constitutional Court. In addition, the constitution now provides that the competencies of the president may be extended by ordinary legislation, which could easily become a formula for nearly unlimited powers, as the Venice Commission has observed.
The only remaining checks on the powers accumulated by the incumbent in office are impeachment, regular elections and term limits. But here, too, the devil is in the details. The vote thresholds for parliament to initiate impeachment investigations are so high as to render the process all but inoperable except in the unlikely scenario that an opposition party commands a super-majority of seats in parliament. While the president is ostensibly limited to two five-year terms, the amendments also provide that if the parliament dissolves itself prior to completion of the second term, the president may run again for office. If the president’s party commands a legislative majority, it would not be difficult to imagine gamesmanship enabling the president to remain in office well beyond the ten years contemplated by formal term limits.
The provisions affecting the judiciary, the final category of changes passed in the 2017 referendum, are less extensive than those reallocating powers between the legislative and executive branches, but are perhaps more devastating. On the bright side, the constitutional amendments completed the project of civilianizing the judicial branch by abolishing military courts—including the appellate military system—other than for purposes of internal discipline. Because the Turkish Constitutional Court (TCC) received two of its members from the appellate military courts, the elimination of these courts reduces the size of the TCC from 17 to 15.
The far more consequential reform, however, is the complete overhaul of what had been the High Council of Judges and Prosecutors. From its structure to its name, arrangements for judicial appointments and promotions were revised. The new Council of Judges and Prosecutors (CJP) is reduced in size (from 22 regular members to 13), with four of its members appointed directly by the president. In addition, the minister of justice and the deputy minister are presidential cabinet members that make up two more seats on the CJP, ensuring that nearly half of the council is comprised of presidential appointees (six out of 13). The remaining seven members are appointed by the parliament. None of the members of the CJP are judges appointed by their peers in the judiciary, as is typical democratic practice. Moreover, if, as is likely, the president’s party commands a parliamentary majority, it will be able to control several (or possibly all) of the parliamentary appointees to the CJP, enabling the president to reliably command a majority (if not the totality) of the Council to the detriment of judicial independence. Because the CJP shapes membership in the high appellate courts, which send nominees for appointment to the TCC, the influence of the executive on the constitutional court is also indirectly enhanced by these reforms.
As is evident, “Turkish-style presidentialism” is a system of executive rule virtually free from the constraints of separation of powers. The president has substantial legislative powers, can dissolve the parliament at will, exercises authority over a wide array of domestic and foreign policy matters and wields significant control of judicial appointments and promotions. In the absence of basic institutional checks and balances, the president is accountable only to the ballot box.
Erdoğan’s frequent use of referenda as a proxy for the popular will has deftly inverted the very idea of electoral processes as a meaningful check on power. Presenting complex policy decisions or elaborate changes to the structure of the constitution as a single “yes/no” choice to be put to “the people” has enabled Erdoğan to short circuit deliberation and reduce debate about the country’s political trajectory to a referendum on his own popularity.
A case in point is the referendum by which the constitutional amendments were adopted in April 2017. All of the amendments were put to voters as a single package, requiring a yes/no vote. During the months leading up to the vote, the choice was largely presented in personalistic terms as a vote for or against Erdoğan. The substantive provisions of the amendments themselves were rarely presented in an accessible way to voters. The polling places themselves featured little by way of information about the content of the amendments. The referendum was approached as an exercise in decisionism rather than deliberation.
The winner-takes-all character of constitutional amendment by referendum means that procedural protections to ensure that the vote is fair are especially important. Voting over significant constitutional changes under a state of emergency is problematic at best. The Organization for Security and Cooperation in Europe (OSCE), for one, suggested that diminished individual freedoms and restrictions on civil society and the media under emergency rule compromised the democratic legitimacy of the referendum. The OSCE was also scathing about the pro-government media landscape in the run-up to the vote and irregularities on the day of the referendum itself.
The parliamentary debate on the amendments was held at a time when the government had jailed 11 opposition MPs. Throughout the referendum campaign, dozens of journalists were similarly detained, creating an atmosphere of intimidation that limited the scope of public debate. For weeks before the vote, the country was awash in “yes” campaign materials in support of the constitutional amendment package. The “no” campaign was all but invisible with its ability to hold meetings, display posters and access media coverage severely curtailed. Opposition meetings and rallies were subject to intimidation and repression by authorities across the country and controls over traditional and social media meant that only the “yes” campaign received coverage.
The deeply uneven electoral playing field produced by these measures yielded a contested and exceedingly narrow victory for Erdoğan with 51 percent of the vote (the opposition argued that the referendum was in fact narrowly defeated, but that the government engaged in ballot stuffing to rig the outcome at the margins when it allowed unstamped ballots to be counted). The narrow result avoided the appearance of crude vote manipulation, though for all intents and purposes the vote was unfree and unfair. With political opponents excluded from contestation, rights of speech and assembly curtailed and the independence of the courts, media and civil society abrogated, it is hard to imagine the ballot box serving as a significant constraint on the president.
The approval of the referendum set the course for Turkey’s constitutional transformation, but did not accomplish it directly. While some provisions went into effect immediately—notably, the restructuring of the CJP and permission for the president to join (and lead) a political party—the transition to the executive presidency system would only take effect following combined elections for parliament and the president, originally scheduled for November 2019. Opposition parties vowed that should their candidates win the presidency they would do away with presidentialism and return to parliamentary government.
But fearful that a looming economic crisis might damage the AKP’s electoral performance, Erdoğan called snap elections nearly 18 months early, giving opposition parties only a few weeks to mount their campaigns. The early vote was held on June 24, 2018 with emergency rule still in effect. The election campaign witnessed a surprising degree of contestation, with opposition parties from the center-right to the left campaigning actively against the AKP in the parliamentary elections and fielding presidential candidates.
Despite hopes surrounding the unexpected popularity of opposition presidential candidate Muharrem İnce, Erdoğan secured the presidency. Moreover, Erdoğan’s election resets the clock on term limits; the logic goes that this will be his first term as president under the new constitutional order, enabling him to run again in 2023. A further quirk in the rules on term limits discussed above may even allow him to run a third time. The AKP, for its part, lost some vote share but won a plurality of seats in parliament and joined its traditional coalition partner, the far-right ultra-nationalist National Action Party (MHP), to form a parliamentary majority. Once Erdoğan assumed office following the vote, Turkey’s system of parliamentary government was abolished.
L’etat c’est lui
The comparative politics literature has a catalog of terms to capture amalgamations of majoritarian politics with authoritarian rule like Turkey’s new presidentialism: competitive authoritarianism, illiberal democracy, electoral authoritarianism, hybrid regimes and the like. The comparative constitutional law literature is swiftly catching up, with categories like “populist constitutionalism” and “autocratic legalism” that capture the means by which constitutional tools are deployed to subvert democracies from within. What they describe is methods of democratic reversal that place constitutions at the heart of the project of consolidating power. This is not like old-school authoritarianism marked by coups and strongmen seizing power through coercion.
For the new breed of electoral authoritarians, democratic processes themselves are used to incrementally chip away at democratic norms. Understanding constitutions not as a means of limiting government but as a way of amassing power and disabling checks is fundamental to their ambition. Erdoğan is arguably at the vanguard of this phenomenon, having deployed the entire array of legal and constitutional tools to accomplish democratic reversal over the last five years: from the packing of institutions to employing state audit authorities to punish adversaries to using constitutional referenda to translate narrow electoral margins into durable consolidation of power. With the introduction of “Turkish-style presidentialism,” Erdoğan has perfected the art of the new constitutional authoritarianism, clearing the way to remain in office nearly indefinitely by disabling most mechanisms of democratic public accountability.
The new constitutional order is one in which the specific powers of ministries, vice presidents, councils, directorates and other state institutions remain very much in flux. The administrative organization of the state is now in the control of the presidency alone, enabling Erdoğan to establish by presidential decree the structures to support his system of one-man rule. The transition from the earlier order to the presidential system has involved a chaotic reorganization of the executive branch. In some instances, ministries were merged—as with the combined treasury and finance ministries now headed by Erdoğan’s son-in-law Berat Albayrak. In other cases, new entities connected directly to the presidency, like the nine councils (representing a merger of what had previously been 65 boards, commissions and committees answerable to parliament), are due to gain significant power, but their powers remain unclear months after their creation. Senior bureaucrats have been removed from their positions en masse, whole ministries and offices have been abolished and existing hierarchies within the civil service have been overturned. Neither the new positions that have been created nor the authorities accorded to new appointees have been clearly defined. Few in the newly created senior administrative positions have a clear enough sense of their authorities to sign documents or issue approvals. The only certainty is that they are all answerable only to the president.
The dismantling of the administrative state has ground the business of governing to a near halt. But in the end, the chaos is quite deliberate— the country that innovated the (now global) trend of paranoid conspiracies about the “deep state” has finally done away with its state apparatus. The uncertainty permeating the administrative state is a reflection of the fact that Erdoğan alone controls all the levers and can change the fundamental organization of the executive together with rules, processes, appointments, promotions and tenure for the civil service on a whim. With neither cabinet approval nor parliamentary consultation to cabin his discretion and the new offices and advisers around him serving solely at his pleasure, the system has all the hallmarks of arbitrary rule. So long as the reins are firmly in his hands, the particular authorities of different offices within the executive cannot define the state. L’etat c’est lui.
 See, e.g., Ella George, “Purges and Paranoia,” London Review of Books, May 24, 2018.
 For an excellent detailed survey of these changes, see Serap Yazıcı, “Constitutional Amendments of 2017: Transition to Presidentialism in Turkey,” GlobaLex, October 2017. Another useful overview is Ilayda Güneş, “What’s at Stake in the Turkish Constitutional Amendment Proposal,” International Journal of Constitutional Law Blog, April 14, 2017.
 Kim Scheppele, “Worst Practices and the Transnational Legal Order (or How to Build a Constitutional ‘Democratorship’ in Plain Sight),” (working paper 2016). Scheppele introduced this concept in describing a similar constitutional amalgam in Hungary.
 Opinion on the draft amendments by the Venice Commission on March 13, 2017. See Venice Commission, Opinion No. 875/2017, March 13, 2017.
 Organization for Security and Cooperation in Europe (OSCE), “Turkey, Constitutional Referendum, 16 April 2017: Final Report,” June 22, 2017.
 A further quirk in the rules on term limits, discussed earlier, means he may even be able to remain in office through 2033 (and possibly beyond) if parliament were to dissolve and call early elections before he completes a second term.
 David Landau, “Populist Constitutions,” University of Chicago Law Review, vol. 85 (2018).
 Kim Scheppele, “Autocratic Legalism,” University of Chicago Law Review, vol. 85 (2018).