Changing the rules of the game

On 8 October 2026 the Italian Chamber of Deputies gave final approval to a new electoral law for both chambers of Parliament, by 227 votes to 164 in a secret ballot. The law, informally referred to as the Stabilicum, replaces the system in force since 2017 and will govern the next general election, due by the autumn of 2027.

Electoral reform is a sensitive subject. It concerns the rules under which every party competes, and it is almost always approved by some of those parties and opposed by others. Here I am not interested in evaluating its implications for representative democracy. I try to address a narrower, empirical question that I am discussing this term with the students of my graduate course on Public Opinion and Elections at Luiss: do the parties that change the electoral rules actually gain from them? And does it matter how close to an election the change is made?

What the new law changes

In brief, the new law moves Italy from a mixed system, in which about 37 per cent of seats were allocated in single-member districts, to a proportional system with a majority bonus:

• Proportional allocation in multi-member constituencies, with the abolition of single-member districts.
• A majority bonus of 70 seats in the Chamber and 35 in the Senate for the list or coalition that obtains at least 42 per cent of the vote in both chambers. If no one reaches that threshold in both, seats are distributed proportionally. There is no run-off.
• A ceiling of 220 deputies (55 per cent) and 113 senators (56.5 per cent) for the winner.
• Thresholds of 3 per cent for lists and 10 per cent for coalitions.
• Candidate lists with a blocked first candidate and up to three preference votes for the others.
• The obligation to name a coalition leader when lists are filed, although the name does not appear on the ballot and the President’s prerogatives in appointing the Prime Minister are left untouched.

The reform has been criticised by the opposition parties on several grounds, including the majority bonus, the number of signatures required of new parties, gender balance and the organisation of the vote abroad, and legal challenges aimed at bringing the law before the Constitutional Court have been announced. Next I focus on what we know about the consequences of reforming electoral rules.

A familiar pattern

In the last three decades, Italy has changed its electoral law more often than most established democracies, and it tends to do so close to elections.

Table 1. Electoral laws approved by the Italian Parliament since 1993. The Constitutional Court struck down parts of the Porcellum in 2014 (ruling no. 1/2014) and of the Italicum in 2017 (ruling no. 35/2017).

None of these reforms produced an obvious reward for the parties that promoted them. The Mattarellum was largely forced on Parliament by the 1993 referendum, and the parties then in government collapsed in the 1994 election. The centre-right coalition that approved the Porcellum lost the 2006 election by about 25,000 votes in the Chamber. In 2018, the parties that supported the Rosatellum had very different fortunes: the Democratic Party fell from 25.4 to 18.8 per cent, while the League rose from 4.1 to 17.4 per cent, and the largest party was the Five Star Movement, which had opposed the law.

Four cases are not evidence, however. To go beyond anecdotes, we need comparative data.

The comparative evidence: no systematic advantage

The intuition behind most accounts of electoral reform is straightforward: rules change when the parties that have the power to change them expect to gain seats. If that were the whole story, reformers should do better after a reform than before.

The most systematic test of this expectation for Europe is a study by Veronika Patkós and Árpád Stump, published in Acta Politica. They examine electoral reforms in 30 European democracies between 1960 and 2011 and define as “reformers” the governing parties that approved a change to the electoral law during the legislature. They then look at three outcomes at the following election: the change in the advantage with which a party converts votes into seats, the change in its seat share, and whether it remains in government.

Their results run against the intuition:

1. Reformers do not gain. On average, the effect of having reformed the rules is negative: reformers lose seat share and are less likely to stay in office. Their vote–seat advantage also declines, although this effect disappears once their vote share is taken into account, which suggests that the loss comes mainly through votes rather than through the formula itself.
2. Major and minor reforms do not differ. Reforms that substantially change the system are no more rewarding than minor adjustments.
3. Timing makes no significant difference. Reforms approved less than a year before an election are associated with a slightly worse outcome for reformers, but the effect is not statistically significant.

A more recent study by Gergely Rajnai, published in Politics & Policy, broadens the picture. It covers 1,055 democratic elections in 141 countries between 1974 and 2017 and counts as a reform any change to the electoral rules, including minor ones such as the number of seats, which yields 324 cases. Reformers improve their ratio of seats to votes neither relative to the previous election nor relative to other parties. Unlike Patkós and Stump, however, Rajnai finds no significant penalty either: the effect is simply null. The two studies thus differ on whether reformers lose, but agree that they do not gain.

Why would parties write rules that do not help them? Patkós and Stump point to two explanations. The first is that reformers often pursue other goals, such as governability, agreements among political elites, the representation of minorities or gender balance. The second is incomplete information: the effects of electoral rules depend on how parties and voters adapt to them, and they are notoriously difficult to predict. Rajnai adds that governments are rarely the sole architects of a reform, since courts, opposition parties, experts and public pressure also shape the outcome. Rival parties may also adapt in ways the reformers did not anticipate, for instance when a fragmented camp unites in response to rules designed to penalise fragmentation. The Italian experience fits these explanations. In their analysis of the 2005 reform in Electoral Studies, Alan Renwick, Chris Hanretty and David Hine show that the governing coalition was pursuing power in a broad sense, including its own internal balance and the conditions under which a future government would operate, rather than simply maximising seats. And a recent process-tracing study of the 1993, 2005 and 2017 reforms by Alessandro Di Biagio in Humanities and Social Sciences Communications shows that party incentives, while important, were never sufficient on their own. Referendums, institutional constraints and the Constitutional Court shaped the outcome each time.

Does timing matter?

The question of timing deserves a closer look, because two mechanisms pull in opposite directions.

On the one hand, a reform approved shortly before an election leaves rival parties and voters little time to adapt, which should preserve the intended mechanical effect of the new rules. On the other hand, it maximises the visibility of the reform and makes it easier for voters to attribute responsibility to the government, which may produce a behavioural backlash. The two effects can offset each other: a party may win fewer votes but more seats.

Research on other types of reform shows that timing can matter. In a study of structural economic reforms in 90 countries published in the Journal of the European Economic Association, Alberto Alesina and co-authors find that liberalising reforms are costly for incumbents when implemented close to elections, because their benefits take time to materialise. But economic reforms work through different channels, and that result cannot be transferred to electoral rules.

There is also a methodological difficulty. Timing is not random. Governments may change the rules shortly before an election precisely because they expect to lose. A simple comparison between early and late reforms cannot separate the effect of timing from the expectations that led to it.

What citizens know and think

If any electoral penalty exists, it must pass through voters. Here two pieces of evidence are useful.

The first concerns Italy. An Ipsos poll published in June 2026, before the first parliamentary vote, showed that knowledge of the reform was limited. Only 29 per cent of respondents correctly identified the parties proposing it, 53 per cent did not know whether a proposal had been presented at all, and only 6 per cent said they had looked into its content. Opinions were divided and followed partisan lines. The majority bonus was supported by 59 per cent of government voters and opposed by 58 per cent of progressive voters. Overall, 36 per cent were favourable and 40 per cent critical. Two procedural questions are particularly relevant here: 47 per cent disapproved of a reform passed without the opposition (21 per cent saw no problem), and 40 per cent considered it wrong to change the electoral law about a year before an election (28 per cent saw no problem).

The second comes from Hungary, where in 2014 the governing party, holding a two-thirds majority, rewrote the constitution and the electoral law before the election. In a survey experiment published in the Journal of Comparative Economics, John Ahlquist, Nahomi Ichino, Jason Wittenberg and Daniel Ziblatt find that informing respondents about the changes did not, on its own, alter their views of the legitimacy of the election. When the information came with partisan cues, assessments became more negative, but only among those who did not support the governing party. Differences between partisan groups were far larger than any effect of information.

Taken together, these findings suggest that electoral rules are a low-information issue on which citizens rely heavily on partisan cues. Concerns about how and when a reform is approved are widespread, but they are concentrated among those who already oppose the government. That may help explain why the comparative data show no systematic advantage for reformers, but also no clear penalty linked to timing.

What to expect

Three conclusions emerge from the evidence.

1. Changing the rules does not guarantee an advantage. In Europe, parties that reform the electoral law have on average done no better, and often worse, at the following election. Worldwide, they do not obtain a more favourable conversion of votes into seats.
2. The effect of timing is still an open question. The comparative evidence shows no significant effect, but timing is difficult to study because governments choose it strategically.
3. Citizens judge electoral rules through partisan lenses. Knowledge is limited and opinions follow the divide between government and opposition.

The 2027 election will offer a new case, under rules that have yet to be tested and that may still be reviewed by the Constitutional Court. As with any single case, it will not settle the question. But it will allow us to observe whether the mechanical effects of the new law, the reactions of voters and the strategic adaptation of parties reinforce or offset one another, which is what this literature suggests we should be looking for.

This post draws on a lecture prepared for the graduate course Public Opinion and Elections (Opinione Pubblica ed Elezioni) at Luiss Guido Carli. Information on the content and approval of the law is based on press reports from October 2026 and refers to the text approved on 8 October 2026.