The Leopard Paradox?

“Everything must change so that everything can stay the same.” (Giuseppe Tomasi di Lampedusa, The Leopard, 1958)

In early May, Prime Minister Giorgia Meloni hosted a roundtable to discuss institutional reforms designed to improve “the stability of governments and legislatures, and respect for citizens’ votes at the ballot box.” A central campaign promise of hers, the reforms are meant to address Italians’ exacerbating distrust of political institutions, rooted in the fact that Italy’s administrations are among the most short-lived in Europe. This adds to its comparatively low levels of ‘clarity of responsibility,’ i.e., the public’s inability to punish (or reward) political leaders. To show her commitment, Meloni nominated former Senate President Maria Casellati as ‘Minister for Institutional Reforms,’ a position created in the late 1980s and only rarely filled since.

Three options emerged from the discussion: French-style semi-presidentialism (“presidenzialismo alla francese”), a German-style chancellorship (“cancellierato”), which includes the institution of ‘constructive no-confidence,’ and the direct election of the prime minister, the so-called “mayor of Italy” (“sindaco d’Italia”) with the abolition of ‘perfect bicameralism.’ I will briefly discuss the potential and challenges of each option.

French-style semi-presidentialism

Championed by the ruling center-right coalition, this proposal intends to strengthen the executive by politicizing the role of the president. Currently, their role is largely ceremonial, though they are formally in charge of appointing the prime minister and the other cabinet members (art. 92) and retain the prerogative to dissolve parliament (art. 88). The coalition’s sponsorship of this proposal follows the tumultuous election of incumbent President Sergio Mattarella in January 2022, who defeated Carlo Nordio – Meloni’s preferred candidate. Electors failed to reach the required absolute majority seven times, demonstrating the limitations of the current indirect election process for the presidency (Italian Constitution, arts. 83 and 85).

The two core features of French-style semi-presidentialism would be directly electing the President by popular vote and granting them a more active role in controlling the parliamentary majority. This could boost stability by allowing presidents to remain a constant when cabinets are reshuffled while furthering respect for citizens’ votes by decreasing the frequency of extra-parliamentary prime ministers (independent or technocratic), a recurring problem since the 1990s. Moreover, it could reverse the growing tendency to elect independent presidents (Ciampi 1999-2006, Napolitano 2006-2015, Mattarella 2015-), instead producing one affiliated with the ruling majority. This could further stability by creating continuity between the legislative majority in parliament and the office of the president, especially in light of findings that copartisan “presidents with significant dissolution powers are able to shape the electoral success of incumbents” (p. 730).

During the drafting of the Constitution in 1946-47, delegates debated and ultimately rejected the possibility of directly electing the head of state, favoring instead an indirect election to ensure the president’s neutrality and impartiality. The choice was driven by the fact that after twenty years of “legally authorized executive dictatorship” (p. 212), as the constituent assembly debates show, fear of personalism reigned supreme. With rare exceptions (e.g. Nepal), no modern constitution has since contemplated an indirect election for the head of state. At the same time, to enhance stability, any electoral reform for the Presidency would also have to involve changes to their powers. Otherwise, Italy would merely approximate countries like Ireland, where, despite the direct election, the President’s role remains ceremonial.

To successfully mimic the French system, some changes are necessary. The 1958 French Constitution contains procedures intended to “rationalize parliament” to achieve a balance between elements of presidentialism and the confidence-based relationship between the government and the legislature, particularly if ‘cohabitation’ occurs (i.e., the President’s party differs from that of the parliamentary majority). These procedures are the ‘package vote’ (art. 44) and the ‘vote of confidence’ (art. 49), “that “artificially ensure” executive decisiveness in the absence of coherent partisan majorities” (p. 2). A similar effect can be achieved in Italy via the ‘confidence question’ procedure, though this is currently not codified in the Italian Constitution. France also eliminated the ex-post vote of investiture to sanction the completion of the government formation process when transitioning to the V Republic, something that is still required by the Italian Constitution (art. 94). Its elimination would allow the head of state to play a heavier role in the formation process by removing the need for explicit support from simple majorities. Finally, France shortened the President’s mandate to coincide with the constitutional term of parliament, something that is currently not the case in Italy, where the President is elected every seven years (arts. 83 and 85) and Parliament every five. Shortening the president’s mandate might lead the same political majority to control different branches of government because the public would vote simultaneously to elect legislators and the head of state, reducing the likelihood of divided government.

German-style ‘chancellorship’

Favored by the Democratic Party, the core innovation of the German-style ‘chancellorship’ option is the institution of ‘constructive no-confidence.’ If successful, a regular no-confidence motion results in the removal of the cabinet, an early election is usually called, or, more rarely, an opposition leader is appointed instead as prime minister. A ‘constructive’ no-confidence motion implies that the legislature votes to withdraw confidence in the incumbent while simultaneously designating a new prime minister. It thus combines two procedures: the no-confidence motion and the ex-ante vote of investiture.

Featured in various constitutions, including Germany, Spain, and Poland, and previously proposed in Italy, including during the Constitution’s drafting, this mechanism could enhance stability by diminishing the uncertainty that surrounds governmental crises. If there is enough support for a new prime minister, constructive no-confidence eliminates the need to call an early election to address the crisis while speeding up the process of replacing the outgoing government. It would also further respect for citizens’ votes by effectively closing off the scenario of extra-parliamentary leaders, as the successor would come from the parliamentary group(s) who initiated and carried the no-confidence motion.

However, to reach these objectives, the procedure must be coupled with institutions that lubricate governance in the presence of non-homogenous social strata (such as federalism in Germany) and regulate the party system more rigidly. In Italy, the opposition has never succeeded in removing the government, though between 1963 and 2006, there have been at least twelve attempts. This failure can be traced to uncertainty regarding the ousted government’s successors, a setting that accommodates fragmentation in parliament, and a dearth of party discipline. This renders it difficult for legislators to coordinate and coalesce behind a new executive. Notably, on its own, the constructive no-confidence procedure would only remove the uncertainty hurdle as assembly members would vote while being aware of what comes next.

A move towards German-style chancellorship also raises questions regarding the appointment process for the prime minister after an election (or any other case that does not imply no-confidence). The 1949 German Constitution prescribes an ex-ante vote of investiture for selecting a chancellor in situations other than a successful constructive no-confidence motion (art. 63), as was reiterated by the German Constitutional Court in 1983 following the Vertrauensfrage case against Chancellor Helmut Kohl. Currently, in Italy, the government must face an ex-post vote of investiture every time a new cabinet is formed (art. 94). Hence, should German-style chancellorship be adopted in Italy, four scenarios open up:

(1) Constructive no-confidence and ex-ante investiture for all other cases, like in Germany.

(2) Constructive no-confidence and ex-post investiture for all other cases.

(3) Constructive no-confidence and ex-post investiture together. This would imply the presence of double investiture requirements, like in Ireland.

(4) Constructive no-confidence and no investiture for all other cases.

It is also unclear whether the president would be constitutionally bound to appoint a leader selected via any of these options.

This proposal would also require the adoption of mechanisms that counterbalance the constructive no-confidence procedure. The German Constitution affords no-confidence prerogatives to the Bundestag (art. 67) but likewise provides the government with the vote of confidence procedure (art. 68). Currently, a critical difference between the German vote of confidence and the corresponding Italian institution (confidence question) is the consequences of failure. In Germany, the denial of confidence upon invocation of article 68 enables the chancellor to submit a request for parliamentary dissolution to the Federal President. While this expedient is not directly mentioned in the Constitution, it has been used three times (1972, 1982, 2005), and the German Supreme Court upheld it twice (1983 and 2005). A similar procedure would grant Italian prime ministers the power to dissolve the legislature, which is currently an exclusive prerogative of the president. The German President’s hands are largely tied when it comes to dissolution (arts. 63 and 68), and it remains unclear if the Italian President’s powers would be restricted accordingly.

Direct election of the prime minister and abolition of ‘perfect bicameralism’

The third option, supported by Azione – Italia Viva, is the direct election of the head of government and the abolition of ‘perfect bicameralism.’ Italy (together with Romania) is an idiosyncratic case in that the Senate is directly elected, enjoys confidence prerogatives, and carries out virtually the same legislative functions as the other chamber.

Holding a separate ballot for the prime minister was tested in Israel in 1996, 1999, and 2001 and labeled a “failed experiment,” which “adds to the risk of a populist or even a plebiscitarian-authoritarian turn” (p. 121). Another concern pertains to reconciling the confidence-based relationship between the government and the legislature and the ‘separate mandate’ given to the PM by the electorate. The latter’s direct election generates tension between the basic idea that parliamentary administrations answer to members of parliament and the ‘legitimacy’ provided to the directly-elected prime minister by the public. Additionally, just as with directly electing the president, it is unclear if or how the prime minister’s and the president’s powers would change or if the head of state would be constitutionally bound to appoint this prime minister. While this option likely aids the goal of respecting citizens’ votes, it is unlikely to increase stability.

The abolition of ‘perfect bicameralism’ can be achieved either by an outright transition to unicameralism or alternatively by reducing or modifying the Senate’s powers. Depending on the final arrangement, either could help reduce instability and expedite the legislative process. Granting confidence prerogatives to only the Chamber of Deputies would cut the number of votes that might result in withdrawal in half, as, under the current arrangements, Italian administrations usually present the same confidence questions to both chambers. This could potentially increase stability because the government would effectively only need to command the confidence of the lower chamber. As a matter of fact, Italy has the only case of a Prime Minister losing a vote of confidence in the Senate I am aware of (Romano Prodi in 2008). Moreover, reducing the legislative capacity of one chamber could increase efficiency and speed up law-making without relying on emergency procedures. Particularly, the lengthy approval process for legislation proposed by the government – an average of 333 days – has pushed Italian incumbents to rely increasingly on decree-laws. This emergency procedure sets a deadline of sixty days for the legislature to amend, ratify, or reject the decree (art. 77).

Taking stock

While all options would change Italian democracy, the least radical and most compatible with the current institutional setting is German-style chancellorship. It is also the only option that does not introduce additional voting responsibilities. Yet, all three have elements that might lead to increased stability of executives and legislatures. The French IV Republic experience shows that the adoption of semi-presidentialism significantly reduced alternations in power. The second proposal would give the legislature a more robust resolution mechanism by replacing the current no-confidence motion (art. 94) with a constructive one. This might ameliorate the recourse to untimely parliamentary dissolutions at the president’s discretion (nine in total since 1947) and reduce the frequency of non-partisan prime ministers (which would also be the case under the French model). Neither the semi-presidentialism option nor the chancellorship one has alarming shortcomings per se, other than requiring majorly reshaping the Italian Constitution (which is notoriously ‘rigid’ and hard to amend). The final proposal is the most troubling. As well as possibly increasing personalism and encouraging dangerous plebiscitarian tendencies, this option is simply not reconcilable with parliamentarism. Reforming perfect bicameralism, on the other hand, might streamline the legislative process and lessen opportunities for confidence withdrawal.