Executive structure reform in Italy?

There are proposals afoot in Italy to depart from the parliamentary form of government, most likely replacing it with some type of semi-presidentialism. In addition, there is discussion of adopting a “constructive” vote of no confidence. (In Italian, see Repubblica, Libre Quotidiano).

Under a semi-presidential executive structure, the head of state (president) is elected popularly, and there is also a prime minister as head of government. The prime minister and cabinet are collectively responsible to the assembly majority. Under a constructive vote of no confidence, the majority that votes no confidence must also name a replacement prime minister. The two provisions are not often combined, although Poland has a semi-presidential system with a constructive vote (see Art. 158 of the Polish constitution).

The Brothers of Italy, party of the current prime minister, Giorgia Meloni, had in their manifesto for the last election a pledge to change to “direct election of the president.” One might presume that the subtype of semi-presidentialsm would be premier-presidential. Under premier-presidentialism, the president may have various initiative rights in proposing a premier following an election or resignation of incumbent premier, but does not have constitutionally delineated power to dismiss the premier, cabinet, or ministers.

A premier-presidential model certainly fits better with a constructive vote, given that both institutional features emphasize the primacy of the parliamentary majority in determining who is premier, in case of conflict. Nevertheless, the Repubblica article (linked above) states that the current proposal calls for “A President of the Republic …who presides over the Council of Ministers and can dismiss ministers.” The inclusion of power to dismiss would imply the other subtype of semi-presidential: in a president-parliamentaty model, the cabinet must maintain the confidence of both the parliamentary majority and the president. President-parliamentarism is, in general, a recipe for instability and competing legitimacy. When combined with a constructive vote, it would imply that a parliamentary majority could say “we want this person to be premier” and the president could turn around and dismiss him or her. Let’s hope when the proponents of this reform say that they want a “French” model they actually mean it. France has a premier-presidential system.

The debate could get even more interesting. There is a statement quoted by Libre Quotidiano (linked above) from Senator Carlo Calenda (Azione Party) and Matteo Renzi (former prime minister) in which they say they do not favor election of the president, but would be favorable to direct election of the prime minister (citing the example of Italian mayors). See also Agenzia Nova where Calenda expresses support for this idea along with a unicameral parliament. The point on unicameralism is important. Italy currently is a rare case of parliamentarism in which the cabinet can be voted out by either of the two chambers of a bicameral parliament. Having a cabinet that has to keep a president and two chambers supporting its continued tenure is probably unwise.

The proposal for semi-presidentialism also calls for changing the term of the head of state to five years. The current unelected presidency has a seven year term.

Italian reform debates will be worth keeping an eye on.

(Thanks to Francesco Bromo for sharing these links with me.)

26 thoughts on “Executive structure reform in Italy?

  1. Unless they have changed things recently, Italian mayors are elected with a guaranteed three-fifths majority on the city council. That doesn’t sound like a reform for stable democracy if implemented at the national level.

    On the other hand, I could see Italian governance improving with a directly elected president who can appoint a cabinet that stays unless an opposing majority exists who can force their choices into office.


  2. The bill being referred to in the Republica article that was rejected in the prior Parliament is this one, I think (http://documenti.camera.it/leg18/pdl/pdf/leg.18.pdl.camera.716.18PDL0015210.pdf). With regard to JD’s question, from my reading of the Google Translate version the answer is that a VoNC from either chamber naming a designated successor for PM is enough to remove the incumbent and replace them with said designated successor. That seems like an unusual solution to me, and not quite “constructive”, given that the PM needs an investiture vote in both chambers.

    Another note from the bill that will no doubt be of tremendous interest to Tom and Alan is that the formal title of “President of the Council of Ministers” is replaced by “Prime Minister”.


    • Oh, great, so Italy may be inventing a whole new form of no-confidence vote. Maybe we should call it semi-constructive.

      But why would there need to be investiture as well? A normal constructive vote is basically a no-confidence vote and investiture rolled into one. It would seem if you have “perfect” bicameralism on Italy’s model, you should choose either: (1) both houses must pass, by majority, a resolution replacing the PM with a new one, or (2) a majority in favor of a new PM in one house is sufficient to invest the new government.

      Obviously, I am not recommending option 2. It could result in ping-pong no-confidence measures, and there would be nothing “constructive” about that. That is at least as bad as president-parliamentarism, which I hope is not what is being proposed.

      At least they are getting the name of the post correct!


      • I think JD is probably right that there isn’t a totally straightforward way to combine constructive votes of no confidence and perfect bicameralism. The proposal in the bill wouldn’t be quite constructive, given that a PM could be turfed from office and their replacement, nominated by one house, could be rejected by the other. On the other hand, neither of your proposals seem to quite be “perfect bicameralism” – the first one lets a Prime Minister remain in office even if one house votes to remove them, while the other lets a Prime Minister ascend to office with the risk of an immediate no-confidence motion by the other house.


  3. The distincton between premier-presidential and president-parliamentary types of semipresidentialism is formally very clear and often useful, but in practice it’s only a difference in the extent to which the president dominates the majority (in government as well as in parliament) and I find that the formal classification obscures this fact a bit. That is why in France, for instance, which is today a premier-presidential regime, but was a president-parliamentary regime previously, the president is still able to dominate the political game despite reforms curving the great powers he enjoyed before, particularly those in relation to the legislative (rather than the very broad exceptional powers). The double executive power therefore remains the key feature of presidentialism and all subtypes are in fact just dots along a line marking the extent to which power is wielded by one or the other executive component. This is proven by the fact that the political dynamics between president and premier in both cases – premier-presidential subtype and president-parliamentary subtype of regime – usually involve a new majority within the executive structure or the loss of a majority by one component that can lead to a stalemate (cohabitation) within the executive structure. Of course, the process is complicate by the type of relation that exist between parliament and executives, for instance the very frequent case of executives run by non-parliamentary premiers, but in effect there is only one semipresidential regime, either with a presidential majority or without one.


    • I agree that the premier-presidential vs. president-parliamentary distinction matters a lot less when the president’s party has a majority (it still matters a lot when s/he does not).

      When was France (formally) president-parliamentary? To my knowledge, the president never had formal power to dismiss the prime minister.


      • Samuels and Shugart would agree with what Bogdan writes in the first sentence (through the parenthetical clause) as well! The classification does not in any way obscure this point.

        France has been premier-presidential since the adoption of direct election of the president in 1965 under the Fifth Republic. It is the case that proves the point that my work (and also that of Elgie and others) emphasizes: Presidents in France effectively can turn the premier into their own agent when they are the head of the assembly majority, but not otherwise. If it is president-parliamentary, by definition they have the constitutional authority to remove a premier even when the latter is the preferred choice of the assembly majority.

        Also “cohabitation” is not equivalent to stalemate.


  4. What about a directly elected President that is also the Prime Minister, but isn’t Prime Minister in the situation where the President’s Party doesn’t have a majority? Does any semi presidential countries have anything like this? An investure to elect the President as Prime Minister, and a vote of no confidence demotes the President stripping him of the powers of Prime Minister. Would something like this work?


    • If the directly elected executive official is also the premier, it is not a semi-presidential system.

      I am not sure I understand the rest of the questions.


  5. The linked bill allows the President to freely dissolve the Parliament at any time, except during the first year after the parliamentary election, and it does not change the Parliament’s power to impeach and remove the President by an absolute majority in a joint session (art. 90). Sounds like a recipe for a Peru-style Linzian nightmare.
    It also makes the President the head of the Government, and directly responsible (to whom?) for the Government’s policy, even though cohabitation is theoretically possible with the constructive vote of no confidence. That’s a recipe for massive presidential-prime ministerial conflict in case of cohabitation (or even outside of cohabitation).
    It looks like the goal of the reform is presidential executive dominance, which would be awful in itself, but the actual result might as well be far greater constitutional chaos and institutional conflict. If this passes, I’m afraid Italy is going to miss proportional parliamentarism.

    Liked by 2 people

    • Although on a second look, the bill does provide for the Parliament to be elected two months after the President, which should eliminate the possibility of cohabitation. It seems they’ve read Prof. Shugart’s posts on the honeymoon effect. Looks like French-style imperial presidency it is.


  6. Article 12 of the 1958 Constitution of the 5th Repubic gives the French president the power to disolve the legislative. That’s enough to make France a presidential-parliamentary regime. Article 8 gives it the power to confirm and dismiss ministers, article 9 the power to chair government sessions, article 10 to veto legislative bills and article 11 to organise a referendum on government initiatives. The distinction between premier-presidential and president-parliamentary subtypes of semipresidentialism is empirically drawn using a scalar of executive, legislative and even other informal powers of the president (e.g. Roper). It is true that article 21 of the French constitution explicitly states that the prime-minister directs the actions of the government (and the prime minister is responsible to parliament), but in practice it is subserviant to the president. It still is.


    • Bogdan, that is incorrect. There is nothing in the standard definition of premier-presidential system that says dissolution power puts a case outside that subtype. You are, quite simply, misunderstanding my work, as well as Elgie’s, and that of many other scholars working in this area.

      Veto power also does not render a case outside the subtype (see Poland), but I am aware of no literature that codes the French presidency as having an actual legislative veto.


      • I have not based my view on the stronger than usually taken into account role of the French president on the veto power of bills (I also should have said the power to confirm or discharge ministers with regard to Article 8 of the French Constitution, not of dissimisal, a dormant power although this sometimes becomes an active power of refusal of particular nominees). The typology is clear, but it sometimes gives the wrong impression that the archetypal French semipresidential president is weaker than other semipresidential presidents, which is not how things most people perceive the political reality. A president with the power the disolve the legislative is in a far stronger position to bargain a government than one without that power : it’s all I am saying…This can mean that the government isn’t simply the reflection of the legislative, but of an implicit bargain between president and legislative…therefore more presidential-parliamentary rather than premier-presidential…


        • If a president faces an incoming assembly controlled by the opposition, his dissolution power isn’t worth anything until the opposition becomes less popular. When the opposition becomes less popular (and the president’s party more), there’s no reason for him to bargain – he can just call an election. Reality is a little more complex, obviously, but what I just wrote describes the French episodes of cohabitation very well, I think. When an unfriendly assembly was elected, presidents didn’t try to bargain for a better government – they simply appointed the opposition leader as PM. As soon as the electoral fortunes has turned around, they called an election – and subsequently appointed their own choice as PM.

          Liked by 1 person

      • The current Finnish president – folowing the three succesive constitutional reforms since the fall of the USSR – does not have the power to remove the prime-minister, yet his dissolution powers of the legislative are worth as much. Finland was always classified as a semipresidential regime of non-parliamentary nature, being one of two which is based historically on explicitly non-parliamentary premises, a sort of republican monarch – absent the practical institution of a monarchy as initially voted in 1918 – with great reserve powers governing with parliament but not responsible to it (the other one is the Austrian case, which takes inspiration from the US and institutes, following many debates before the fall of the Empire, a republican federation with a collegial government in theory subordinated to the president and only to the president, but in practice to parliament). A multi-party legislative expands the scope of the president dissolution prerogative, which is also the case of France, to the point that the now common place, even traditional, distinction between a premier-parliament style of semipresidentialism and a president-parliamentary style of semipresidentialism is blurred. Without the presidential dissolution power, the distribution of power between president, premier and parliament changes significantly in a premier-president type of semipresidentialism. I surmise that the power of the premier becomes autonomous, being capable of maintaing itself independent of the president (as parlamentarism would require), but sometimes even independent of parliament, in the sense of independent of a divided majority (as presidentialism requires it).


        • A further example of the fragility of current classifications : Romania – where every limit of the regime has been fought over during the last two decades or more – is customary classified as a premier-presidential subtype of semipresidential, at least since the 2003 constitutional revision which formally eliminates the rather unstated power of the president to remove the premier that was used once, in 1998, on state of emergency grounds more than on constitutional grounds even then. Nevertheless, an influential figure in the development of the theory like Sartori dissents that it is even a semipresidential regime, although he recognizes some elements “from the French Fifth Republic”, such as direct election and the power to consult the people by referendum (the actual sources of the 1991-1992 constitutional framers are more ecclectic, but a bit unclear). What has emergence from the president-premier-parliament disputes of the last decades that have shaped Romania’s internal politics is more in line with Sartori’s suggestion of a parliamentary rather than a semipresidential regime. The Romanian jurists and the politicians insisted on one might call “the mediator clause”, i.e. a constitutional paragraph defining the Romanian president as “a mediator of institutions”, not an executive coordinator like the French one, whose origin dates back to the pre-comunist parliamentary constitutional monarchy. Nevertheless, I would not say that Romania is a parliamentary regime with a directly elected president, but neither would I say that the semipresidential regime in Romania today is practically equivalent to the one in France as the subtype premier-president semipresidentialism in which both these countries are grouped would imply.


        • Erata : The Romanian president revoked the premier once in 1999, during the 1998-1999 crisis.


  7. The “obscures a bit part” is based on practical or anecdotical observation, simple conversations : people tend to focus on the prerogatives rather than on the interaction between the institutional actors. This being said, I still view France as rather presidential-parliamentarian than premier-presidential no matter what a given score says. I understand the reasons behind the classification, that’s why I said it is formal and descriptive. I normally use a more simple, maybe more imprecise, even antiquated, but quite effective classification that emphasizes that it’s all a matter of degree : strong and weak semi-presidentialism. The French political regime is, for sure, not the Russian political regime. Nevertheless, in practice too often a French president gets his own premier without actually triggering an election, mearly by threatening to do so; in any case, it dominates the legislative parties. More recent reforms have bolstered the French legislative, some would even say turned it into an actual legislative : the two chambers can now have joint sessions in some cases and the Assembly can actually set its own work agenda, which is almost tantamount to having legislative initiative (usually a defining trait of a legislative body!). But the direct election of the French president has more often than not bolstered the president not weaken, because one way he influences the whole political processs beyond the explicit prerogratives is through the electoral effect the presidential elections have over the legislative elections. The case of Emmanuel Macron is paramount in this regard : he created a whole presidential party from scatch due to this effect (almost like De Gaulle at the outset of the Fifth Republic)! True, cohabitation is not always a stalement, particularly when you have governing coalitions and the president – like D’Estaing – comes from a secondary party (thing were different with Chirac and Jospin!). The statement must be influenced by the observation of a different semipresidential regime in which more often than not things do turn into a stalement during cohabitations. It depends on the party system, I think : if it acts like a bipartisan system the probability of a stalement during cohabitation is very high; if it is more fragmented ideologically the probability of stalement is lower.


    • The issue is that a classification based on real power relations rather than the on the provisions of the Constitution will be unavoidably much more subjective.
      This is why Shugart, Carey, Samuels, Elgie etc adopted the more objective classification of premier-presidential and president-parliamentary and dropped Duverger’s requirement that the President have “substantial powers” for a regime to be semi-presidential. The classification is usually a useful heuristic for how powerful a President is going to be, but these are two different things and shouldn’t be confused, as the examples of the President of France and of e.g. Austria make clear.
      If you want a classification based on actual presidential powers, it’s probably better to use for instance Siaroff’s classification (symbolic, corrective and dominant presidencies) rather than try to change the meaning of the premier-presidential/president-parliamentary classification.


      • No, the classification on constitutional provisions is the more objective one. Only sometimes are the relevant constitutional provisions ambiguous. “Real” powers classifications are certainly more subjective, even though they can be done, and my work also does them.

        Both classifications on institutions and classifications on behavior have their place and value. However, one should not redefine existing institutional classifications to be behavioral ones, or else one is just obscuring the topic. We have an established subtypes based on institutional variation. There is no question whatsoever that in this framework, France is premier-presidential and has been since 1965. Because I have not seen the full relevant text I do not know yet which subtype is being proposed for Italy, which was meant to be the main point of this thread.

        We don’t seem to be getting anywhere with this little “debate” so perhaps it is not worth continuing.


      • You make a good point. But I haven’t disputed the core assumption regarding the preeminence of constitutional provisions (legislative and executive) in the classification of semipresidential regimes in any of my comments – maybe just suggested a more, yes, behaviourist, or interactionist I’d say interpretation of the said provisions. My worry is that you have several counter-intuitive results with its normal use, such as a French president which is as strong or even weaker than a Romanian president, for instance. This might be true, but most everything in following actual politics says the contrary, and people then have to patch the theory with auxilliary assumptions regarding a flaw and the like….There might be a missing subtype or some premises in the theory that are insufficiently well formulated.


      • The classification based on constitutional powers is in line with the tradition of regime theory since Antiquity, since Aristotle. In the Shugart-Casey restatement it is based on the last mover’s advantage, either the president or parliament. This usually predicts the power of the institutional actors, or in any case is in line with the modern political philosophy’s penchant to start with the extreme, crisis or hobbesian, situation at its core – so Duverger’s substantive condition is sort of satified. I think I reason from within the established bounds of the mainstream typology.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.