Italy

The bicameral Italian Parliament has 400 deputies and 200 senators, who are elected every 5 years, both chambers in a mixed system similar to parallel voting. The Senate is also supplemented by a few senators appointed for life.

The minimum age for active suffrage (the right to vote) is 18 years, and for passive suffrage (the right to be elected) is 25 years in case of the Chamber of Deputies and 40 years in the case of the Senate. Voting is not compulsory, and voting is possible through diplomatic missions or by mail (postal voting), but online or proxy voting is not possible.

Chamber of Deputies

147 deputies (37% of the seats) are elected in single-member constituencies and 253 via regional closed-list proportional representation, 8 of the latter from overseas constituencies (overseas residents can only vote in their own list-based constituencies). The election is essentially carried out with a single (fused) vote, where the the voter cannot split their vote between an individual candidate and another party's list (unlike under parallel voting).

In the national list election, the legal threshold is 3%, and the simple quota (Hare quota) and largest remainder method are used to allocate seats between parties.

Senate

74 senators (37% of the seats) are elected in single-member constituencies, and 126 in regional closed-list PR, 4 of the latter from the constituencies of overseas residents (overseas residents can only vote in their own, list-based constituencies). The election is essentially carried out with a single (fused) vote, where the the voter cannot split their vote between an individual candidate and another party's list (unlike under parallel voting).

In the national list election, the legal threshold is 3%, and the simple quota (Hare quota) and largest remainder method are used to allocate seats between parties.

The New Electoral System (2027-): 'Melonellum'/'Stabilicum'

The electoral system reform in Italy is nearing the home stretch; its main feature is the return of the winner’s bonus (jackpot). It would be hard to deny that this looks bad considering the law’s Italian historical precedents.

Over the last 35 years, Italy has changed its the electoral system almost as frequently as their prime ministers. The long-standing system of proportional representation with open lists unfortunately fell victim to reforms in 1993; the disastrous ‘Mattarellum’ was the first to introduce single-member districts and a vote linkage corrective mechanism—similar to Hungary. (Yes, the law was nicknamed after the current President of Italy, who was an MP at the time and worked on the legislation.) A few cycles later, Berlusconi derailed this system and attempted to introduce a winner’s jackpot system; this was known as the ‘Calderoli Law’ or the ‘Porcellum’ (something like “rascality”). After the Constitutional Court struck it down; its remnants lived on as a proportional system under the name ‘Consultellum’. Subsequently, Renzi’s reform efforts sought to introduce yet another jackpot-based system for the Chamber of Deputies known as the ‘Italicum’, but this also was partly struck down by the Constitutional Court. Moreover, the broader reform package failed, leading to yet another reform. This resulted in the current — and, in my opinion, once again terrible — ‘Rosatellum’, a mixed majoritarian (mixed non-proportional) “not-really parallel voting” system that remains in force for now. This is expected to be replaced by the ‘Melonellum’, or: ‘Stabilicum’.

What are the key points of Meloni’s reform?

• Abolition of single-seat constituencies (“ending FPTP”)

• A jackpot model for both houses: a proportional system based on current districts plus a winner’s bonus (jackpot)

• In the Chamber of Deputies, the coalition coalition with the most votes receives a bonus ensuring it holds 56% of the seats, capped at 70 bonus seats (out of a total maximum of 220 seat), provided it secures at least 42% of the vote (and also that it fulfils the same criteria in the Senate). The remaining seats are distributed via proportional representation (nationwide), while the 8 seats for overseas voters are handled separately (which is why I didn’t write 55% as the jackpot figure). In the Senate, the bonus and seat cap (57.5%, or 113 seats, excluding the 4 overseas seats) also apply at the national level, but seat allocation is regionally proportional. Consequently, the Senate’s system introduces a slightly higher degree of distortion.

• The electoral threshold remains at 3% (10% for coalitions)

• Special rules apply to Valle d’Aosta and Trentino-Alto Adige/Südtirol (votes and seats from these regions count towards the majority bonus and the seat cap, but not towards the proportional allocation).

• The number of “overseas” constituencies is also being reduced, though the number of seats remains unchanged. In the Chamber of Deputies, the number of districts is dropping from four to two, and in the Senate, from four to one. Consequently, de facto single-seat constituencies are being eliminated here as well.

Before passing judgement on the system, let us look at the context: the current system, the original reform concepts, and what changed during the process.

Single-seat constituencies

Italy’s electoral systems have been disastrous for over three decades; perhaps the most sensible reform attempt was that of Renzi in 2015–16, though frankly the standard is so low that even Berlusconi’s unconstitutional “shenanigans” of 2005 rank among the better ones. I find the current system appalling — downright malicious toward the voters. The way single-seat constituencies are linked to party lists is, in my view, highly problematic, as it undermines the very essence of individual districts (voters may legitimately except that single-seat district are a way for them to choose their representatives based on an individual basis, not based on their party, if they wish to do so). The fused vote is why I wouldn’t call it a parallel voting system, since there is no option of ticket splitting under penalty of a void ballot, at best it can be called a parallel seat-allocation system (instead of parallel voting, it’s a ‘diverging’ mixed single vote). While the fused/linked vote this might arguably be justified by a desire to foster party/alliance cohesion, the means and the end lack a logical connection. If voters want to make a difference in the local race, they might have to sacrifice their party preference as well. In a single-vote MMP system at least the voter could be sure that ultimately it’s the party vote that matters, but in this system, the local FPTP race may be more important. Meanwhile in uncompetitive districts, voters are free to vote their for their favourite party (possibly making the seats even more uncompetitive). Perhaps one point in the system’s favour is that at least when voting for candidates presented by alliances, the voter may still choose which party they support within the alliance. 64% of seats are allocated via PR, thereby limiting distortions caused by single-seat districts...

Abolishing single-seat districts in favour of a consistent national bonus/jackpot at does serve the goal of ‘governability’ better. While the bonus might be disproportionately large, it still remains a more targeted mechanism than the possibly quite arbitrary distortions inherent in single-seat districts. But, since the bonus operates on a list basis, such a change in itself also eliminates the (kinda fake) personalised voting aspect — a topic to be addressed later in the section on preferential voting. First, however, let us examine the history of winners bonus/jackpot systems in Italy.

Jackpots since Mussolini

In 1923, the infamous Acerbo Law was enacted; under its provisions, the largest party—provided it secured at least 25% of the vote—would automatically receive two-thirds of the parliamentary seats (a constitutional majority). The law was designed to grant Mussolini absolute power in the 1924 election. Yet, even in that election, he circumvented his own rules: since the largest party was capped at receiving “only” two-thirds of the seats (regardless of whether its vote share was 30% or 60%), they fielded a second list to secure an even larger share. Consequently, with a combined vote of 65% (60% plus 5%), he captured nearly 70% of the seats — and they did not bother with further elections after that.

Thirty years later, in the 1953 election, the centrist government coalition attempted to secure a landslide victory using a scheme suspiciously similar to the Acerbo Law. Under the ‘legge truffa’ (“swindle law”), any coalition winning an absolute majority (over 50% of the vote) was entitled to 65% of the seats. They missed this threshold by just 0.2% in the election. The bonus mechanism reappeared in 1993, though not yet at the national level. It was introduced first at the local level for the direct election of mayors, and subsequently at the regional level in a similar manner, although in these cases it involved a separate vote for mayor or regional president. So in practice, this was coupled with ‘directly elected prime minister’ model.

As previously mentioned, Berlusconi reintroduced the jackpot system at the national level in 2005; this guaranteed the largest coalition 54% of the seats — specifically 340 seats, with additional seats available from constituencies abroad. While this system worked to his advantage in 2008, it backfired in 2013 (at least in the Chamber of Deputies; the bonus did not apply nationally for the Senate). The Constitutional Court gutted the system due to its lack of proportionality, effectively leaving behind a proportional system that was never actually used in an election. The subsequent reform was Renzi’s ‘Italicum’, which brought back the jackpot — but only for the Chamber of Deputies — while assigning a completely different role to the Senate (it would have abolished direct elections for that chamber). This time, the jackpot was to be awarded to an individual party rather than a coalition, and in the first round, only if it exceeded a 40% threshold. In the absence of a party list surpassing 40%, a runoff round would have determined the recipient of the jackpot. However, two developments prevented this system from ever being used in an election: the Constitutional Court struck down the jackpot mechanism again — though this time only the one after the runoff provision, as it deemed the bonus for a party exceeding 40% in the first round justifiable. The result was a proportional system featuring this conditional jackpot. However, the rest of the Renzi constitutional reform failed in the meantime, meaning the Senate retained its existing powers while still using the the earlier system. As the situation was considered to be untenable, the entire system was rewritten, resulting in the ‘Rosatellum’ — complete with single-member districts and all. Since then, the only significant change has been a reduction in the size of Parliament, while the proportions between individual constituencies, party lists, and overseas seats have remained the same.

Returning to Meloni’s bill: unlike the ‘Italicum’, this system applies to alliances—much like the ‘Porcellum’ — yet, similar to the ‘Italicum’, it includes a threshold for the majority jackpot. The Constitutional Court accepted the 40% threshold in the previous instance, so, presumably, a 42% threshold (for a marginally larger jackpot) should also be acceptable. Meloni originally wanted a runoff round as well, but that provision was dropped — being precisely that element which had been deemed unconstitutional in the ‘Italicum’ (as it would have allowed a party to secure a massive bonus with less than 40% of the vote). It appears the proposal was carefully calibrated to avoid the fate of its predecessors during judicial review. However, one potential avenue for legal challenge lies in the fact that, unlike the ‘Italicum’, this system involves party alliances, making it more difficult to form single-party governments. Although in practice, the system will likely prove less disproportional than the ‘Italicum’ would have been — and can thus be argued to interfere less with the principle of equal weight of votes — this may also mean it is less effective at delivering on the promise of a “stable government majority”. It could award the bonus to loose party alliances that combine their lists merely for tactical reasons but are ultimately unable to provide a lasting, stable governing majority. Would be ironic if the system ultimately failed because it wasn’t disproportional enough...

Direct election of the prime minister?

Let us also consider the election of a Prime Minister. Meloni and her party originally wanted to move the system towards a semi-presidential system, but since there seems to have been little demand for this, they instead pivoted to the possibility of a direct election for prime minister. (After all, this already exists at the local and regional levels.) The ‘Stabilicum’ seems to contain remnants of this, although not even the alliance that receives the bonus automatically provides a prime minister. However, the parties/alliances, under the threat of deleting the list, still have to nominate a “prime ministerial candidate”. This is also a controversial element, as it is a strange intermediate solution. Neither that prime minister is nominated by the President of the Republic, nor that the representatives have a free mandate (they are not bound to the designated prime ministerial candidate) have been changed. The dilemma, then, is this: if the role of “prime ministerial candidate” carries only political weight, why include it at all? Yet, if it were to hold legal significance too, the aforementioned principles would be compromised. In my view—since the former is the case—the argument that such a requirement imposed on parties is unacceptable is not entirely convincing. Is it a disproportionate sanction to bar a party from running if it fails to do so? Perhaps. But in a parliamentary system—especially in so-called majoritarian, ‘winner-takes-all’ systems — it is entirely natural for an parliamentary election to effectively become a veiled election for the executive. Clinging to the fiction that voters merely elect a parliament, and that MPs subsequently choose a government independently of that process, is a bit absurd. Naturally, there is a universal desire for voters to have the more say in who governs them. Direct election of the prime minister is one model for this (albeit one not without its flaws), but since that did not work out for Meloni, they are left with this ‘lite’ version.

Open list or closed list?

The type of lists is another point of contention. Every Italian electoral system since 1993 has utilised closed lists; voters have only been permitted to vote for individual candidates under the ‘Mattarellum’ and ‘Rosatellum’ systems (though in the case of the ‘Rosatellum’, this was largely nominal) — and even then, only in single-member districts (setting aside the autonomous regions that serve as exceptions). Although the amendment initially failed to pass in the Chamber, the Senate reintroduced the option to cast preferential votes for candidates on the list. The plan allows for three such preferential approvals. The catch is that this does not apply to the lead candidates, and the impact of preferential votes is significantly restricted. According to some estimates, only about 20% of the representatives would actually be elected via preferential votes.

To understand this better, its useful to take a look at how the list-based seat allocation operates. From the voter’s perspective, in a multi-member local constituency, the ballot would display lists of seven candidates; small checkboxes would appear next to six of them (but not the lead candidate), allowing the voter to select up to three candidates. Of course, this is not mandatory; the voter can simply choose the list as a whole. However, lists may appear grouped together with other lists, representing party alliances. If a list is part of such an alliance, the names of the joint list (created to secure the “governability bonus”) will appear beneath the individual list names; no preferential voting is permitted for these joint list candidates. In principle, voters can choose to vote for the alliance itself rather than for individual lists, though doing so effectively delegates the choice of list (within the alliance) to other voters. However, the allocation of seats involves numerous steps before determining how seats are distributed among local lists and candidates. Votes cast for lists are aggregated first at the local level, then at the intermediate regional level, then at the national level, and finally at the level of party alliances. Following the application of thresholds and the potential allocation of a bonus (jackpot), the number of seats to which alliances or non-allied parties are entitled under the proportional component is determined nationally for the Chamber of Deputies and regionally for the Senate. Seats are then distributed among the parties within each alliance; only subsequently are the parties’ seats allocated to regional and then local constituencies. Once the number of seats secured by each list in a district is determined, the local list leader receives the first seat, followed by the other candidates — ranked by the number of preference votes — only if seats remain available for that list. Since districts have a small number of seats (expected to be a maximum of 12–14, but averaging closer to 7–8) and these are shared among many lists, it will frequently happen that only the leader from each list secures a seat, rendering preference votes futile in such cases. That is why the criticism arises that, in this form, preferential voting is merely a gimmick to fool the public. Preference votes are unlikely to grant voters any real say in case of smaller parties; the chance for voters to override the party-determined order will exist primarily with parties that are quite strong locally. When a coalition secures a bonus, fewer candidates can enter parliament via preference votes (as bonus seats take away from the pool that would otherwise go to proportional seats), whereas when there is no such bonus, more can do so.

Regardless of this, I consider the concept of guaranteed entry for list leaders to be interesting. It is quite common worldwide for list leaders to be the the best known candidates — which is often precisely why they get that position on a list — and a phenomenon known as “donkey voting” can occur, where voters wishing simply to support the party cast their vote for the name at the very top of the list. It takes significant opposition for a party’s own voters to reject a list leader; usually, this also requires some form of coordination in favour of another candidate. Therefore, it is not certain that guaranteeing them a seat—as is the case in Italy—will have much practical significance. On the other hand, I wonder, what those voters do, who would have simply voted for the list leader? Will they still try to vote for that person, or will they move on to the next name on the ballot (thereby benefiting the second-place candidate on the list)? Or will they give more thought to their choice, or perhaps consciously forgo casting a preference vote altogether? One thing is certain: since there is no checkbox next to the top-ranked candidates, voters are at least not misled, yet the system provides no direct information to the parties either about the popularity of their leaders.

The devilish details

In this already lengthy analysis, I have focused on the electoral system in the strict sense — specifically, the character of representation. I therefore advise the reader to approach my assessment critically; as I am not an expert on Italian politics, I cannot fully illuminate just how diabolical the system may be as a whole within the local political context. I know well from the Hungarian constitutional and public law of the past 16 years that a system of rules in the aggregate can be highly biased, yet composed of elements that may appear neutral, justifiable, and acceptable in isolation. The same was true of the Hungarian electoral system. The Italian system of Meloni may prove similarly problematic.

It seems new system for collecting endorsements is a cause for concern, for instance; it imposes a significantly higher barrier to entry for new political groups, whereas parties already in parliament are exempt from collecting endorsements or require only a quarter of the amount.

The opposition also criticises the gender quotas — or rather, the lack thereof. Although the ‘zipper system’ (alternating genders of candidates) must be followed on local lists, the guaranteed seats for list leaders and the preferential voting method may allow a skewed gender balance, and the jackpot seats are not subject to these requirements.

Another point of contention for many is that a single candidate may appear on the ballot in up to five different districts, thereby increasing their chances of winning a seat. While I do not consider this unreasonable given the logic of the system as a whole (top-down seat allocation, fixed list leaders), the objections are understandable: despite the small size of the electoral districts, parties can guarantee seats for their preferred candidates, meaning these representatives do not really have to compete personally to win support within a specific district.

What do I find problematic?

I do not think it is a good idea to select the winner entitled to the bonus using the same single-choice system used to allocate the PR seats. If there is to be a bonus, there should be a separate vote for it — perhaps using a ranked-choice system.

The seat cap (which makes the bonus a ‘jackpot’, and the system in some sense indeed ‘compensatory’) is understandable and serves as an important safeguard in certain respects; however, the question arises: to what extent can it be circumvented? Since no bonus is awarded below 42%, only political forces polling above 45% could realistically benefit from using decoy lists. An alliance polling 46% could cobble together roughly 56% plus 4% — totalling 60% — of the seats (give or take, as the bonus reduces the seats won by the 3–4% decoy list, yet that list also benefits somewhat from the votes of parties that fail to cross the threshold). Meanwhile, a force polling at 50% would likely secure only around 64% of the seats, even with significant optimisation; in other words, an absolute majority does not automatically translate into a two-thirds majority. A constitutional majority would likely require around 53% of the vote, and even then, there would be significant risk involved.

A major issue regarding the system's constitutionality on the Senate side is that the bonus operates on a national level, just as for the Chamber of Deputies. Given the constitutional requirement for the Senate to be regional in nature — a principle the system upholds for PR seats — I fail to see how this mechanism would comply. I suspect the Constitutional Court will accept the bonus for the Chamber of Deputies but might strike down the Senate version in its current form (ironically, if a regionally proportional Senate were to remain while its power to grant or withhold confidence in the government were to be abolished, this could actually evolve into a quite sensible semi-parliamentary model).

In my view, there are strong arguments that should a "winner's bonus" exist to guarantee the ability to govern, an electoral threshold (be it 3%, less, or more) is unjustified; the remaining seats could simply be allocated on a fully proportional basis. I do not know if such a provision would be challenged before the Italian Constitutional Court, but I would certainly give it a try...

Concluding thoughts

It is tempting to interpret the ‘Stabilicum’ merely as a means to consolidate the power of the current governing forces. Of course, I do not for a moment believe that Meloni’s legislation is born of pure goodwill; parallels with the direction taken by Orbán in 2011 are certainly justifiable. Yet, the system operates very differently and is specifically designed to keep the winning alliance's seat count below the two-thirds threshold. Looking at the 2022 election figures, Meloni’s alliance secured 59% of the seats with just over 43% of the vote under the current system; under the ‘Stabilicum’, they would have won fewer seats given the same results, due to the seat cap. In fact (without decoy lists), they wouldn't have gained any more even with 45% or 49% vote share.

The primary scenario where the new system is distinctly advantageous for the governing coalition is in case of a victory in a tight race (with a result above 42%), as it could significantly amplify their win; however, this hold if another alliance would win in such a way. Conversely, the ‘Stabilicum’ is not at all favourable for Meloni’s camp if they fall below 42%; in that case, it would be actively detrimental to them, even if they were to finish well ahead of the runner-up, much like in 2022. All in all, the system does not strike me as being as bad as it could have been, or as bad as I had anticipated. Subjectively, I would rank the systems seen so far as follows (bearing in mind that potential translation errors or a lack of high-quality secondary sources might have led me to misinterpret aspects of the text passed by the Senate):

1. Pre-1993 system

2. Gutted ‘Porcellum’ (i.e., ‘Consultellum’, 2014–2015)

3. ‘Melonellum’ / ‘Stabilicum’

4. Partially struck-down ‘Italicum’ (2017)

5. ‘Italicum’ (2015–2017)

6. ‘Porcellum’ (2005–2014)

7. ‘Rosatellum’ (2017–2026)

8. ‘Mattarellum’ (1993–2005)

To explain a bit: proportionality carries significant weight for me, and I also favour open-list systems that offer the greatest possible freedom of choice. However, I consider it equally important that, if a system deviates from these principles to achieve a specific objective, it does so using a mechanism that is coherent, logically aligned with that objective, and reasonably effective. Furthermore, it is crucial to minimise the risk of misleading voters, to ensure that voting can be as sincere as possible, and to limit the scope for organised manipulation of the system. In light of this, I currently view the ‘Melonellum’ as the best Italian electoral system adopted by parliament (rather than one created by the Constitutional Court) in the last 30 years. I make this assessment by examining the system—unlike its architects—from behind a "veil of ignorance"; that is, I do not consider whom it might happen to favour.

The president (head of state) is elected indirectly. More soon.

MEPs from Italy are elected every 5 years (at the same time as in the other EU member states). The country is split up into five (multi-seat) constituencies, in which a total 76 MEPs were elected in 2024. Due to "degressive proportionality", Italian voters have fewer representatives per capita than the EU average.

The main principle in the election of the Parliament (within the countries) is proportionality: by default, only a proportional system can be used, but beyond this, the member states decide for themselves what type. Italy, like most member states, uses an open-list proportional system. This means that voters can vote not only for a party list, but also for a candidate on the list, thus allowing them the chance to collectively change the predetermined order of the candidates. The allocation of seats between parties is done using the simple quota (Hare quota) and the largest remainder method. The legal electoral threshold is 4%, but in some constituencies more is needed to win a seat. 

The minimum age for active suffrage (the right to vote) is 18 years, and for passive suffrage (standing for election) is 25 years (highest in the EU tied with Italy). Voting is not compulsory, voting at embassies is possible (but only from other EU member states), however, postal voting, online voting or proxy voting are not possible. In person voting is possible on two days (Saturday and Sunday). 

This is the description of the electoral system of Italy as of 12.10.2026 on Electoral Knowledgebase. Sources and further information on this topic:

• 2024 European Parliament elections: National rules (European Parlament)

The part about the new system ("Melonellum") is mostly a machine translation of the Hungarian original. The thoughts on it are very much my own, but because of the lack of in-depth secondary sources in English, I have this time relied on ChatGPT to help me understand the mechanics of the system based on the primary source (the text of the law in Italian). I have endeavoured to check its work as much as it seemed reasonable, but for such a complicated system I cannot be completely sure that I have every detail right. The law is relatively new, I will try to keep up on any updates once more scholars comment on it, hopefully, regarding the big picture, I didn’t get anything wrong.

Feedback, criticism is welcome and appreciated.