The Italian Constitutional Referendum: Political and Institutional Consequences of a Striking “NO”

By Gianluca Passarelli

The electoral results of the constitutional referendum have led to the Prime Minister’s resignation. But let us consider what happened before.

On December 4th 2016, Italian voters expressed their vote on a referendum about constitutional reforms. This was the third referendum of its kind in Italy, with the other two held in 2001 and 2006. The two options presented to voters this time were related to the approval or rejection of the reform promoted by Matteo Renzi’s government and his centre-left parliamentary majority. However, several Democratic Party’s MPs decided not to support Renzi’s position, and used the ballot as a tool to oppose their leader due to different visions of the party, the government, policies, and the reform itself. The reform was approved earlier by an absolute majority in both houses of parliament, but the proposed changes required a two-thirds majority in parliament in order to be implemented without a referendum according to the Italian Constitution (art. 138.3). Since this threshold was not met in parliament, the referendum was called (by the Government) by collecting the required number of voter signatures, as stated by the art. 138.2, while the opponents to the reform were not able in getting the minimum number of required signatures (500.000).

The result of the referendum was both clear and decisive. Approximately 60% of voters cast a “NO” vote in opposition to the proposed reforms and only 40% voted in favor. Perhaps the most striking result was voter turnout. Nearly 70% of eligible voters cast a vote, a percentage that is similar to that reached in general elections in Italy (e.g., 75% in 2013). This figure also confirms that Italy remains a democracy with one of the highest electoral participation rates in the world. Despite this high turnout figure, one of the most notable features of the referendum is the persistent North-South divide in terms of turnout and the level of rejection of the reform. Rejection of the referendum was particularly high in southern regions, with peaks in Sicily, Sardinia, and Campania. Support for the referendum was limited and prevailed in only two regions (i.e., Tuscany and Emilia-Romagna), as well as in the province of Bolzano.

A closer investigation of the result reveals a clear centre/periphery electoral pattern, with “NO” support found in less urbanized areas, and “YES” support located in urban and metropolitan areas, though unable to surpass 50%. A positive correlation appears to exist between the support for the referendum and support for the Democratic Party in recent elections. Therefore, the “centre-periphery” cleavage is not necessarily a surprise because the Democratic Party tends to perform better in urban areas. The age of voters was also a factor with younger voters in general more likely to reject the referendum. Interestingly, nearly two-thirds of Italians who participated in the referendum from abroad supported the reform. Since only a few of the smallest parties in Italy and Renzi’s PD supported the reform, the negative result is not entirely surprising though the overwhelming turnout was.

The precise language of the December 4, 2016 referendum was as follows:

«Do you approve the constitutional bill concerning the proposals to overcome perfect bicameralism, to reduce  the number of members of the Parliament, to reduce the operating costs of said institutions, to abolish CNEL and to revise Title V of the 2nd part of the Constitution, which was approved by the Parliament and published in the Gazzetta Ufficiale n. 88, on April 15, 2016?»

What was really at the stake beyond this question? Although the reform modified the text of 46 of the 138 articles of the Italian Constitution, the “real” changes were far fewer and included minor “revisions” as a consequence of double references (here the text of the reform). In fact, the most significant of the proposed reforms were centered around two key issues: eliminating Italy’s ‘symmetric bicameralism’, and reforming the Italian senate.

The elimination of Italy’s ‘perfect’ or better symmetrical bicameralism was central to the reform. Since both the Chamber of Deputies and the Senate maintain an equal share of legislative power, political impasse and obstruction are more common that progress. In fact, as established in 1947, identical texts of Italian laws must be approved by both branches of parliament. Moreover, since both houses confer the confidence to the government, each can autonomously provoke a government’s downfall by withdrawing its confidence. Such kind of bicameralism is unique among contemporary democracies and is a subject of much debate by politicians and scholars alike. It also contributes to political instability especially after 1994, as the risk of different majorities in the two branches has increased due to differences in how and who selects deputies (e.g., minimum voter age is 18 years), and senators (e.g., minimum voter age is 25 years) These different electorates have divergent electoral behaviors that are further accentuated by the regional allocation of seats for the Senate versus the national allocation for the Chamber. Therefore, the electoral geography of Italian politics plays an important role in the allocation of parliamentary seats with increased party and voter volatility.

The collapse of the pre- 1989 party system opened the door to more opinion-based electoral behaviors, together with new parties that were not linked to historical political traditions. The success of the Five Star Movement in 2013, attractive to younger voters (about 44% according to ITANES), and the Northern League since 1992-1994, that is/was especially settled in northern regions, highlight the need for parties to focus on a few regions where the number of seats allocated to the Senate was bigger, as in Lombardy, Campania, or Veneto. This was particularly true for the Senate and especially after 2005. According to those campaigning in favor of the referendum, abandoning the system in which both chambers have equal powers, not only in terms of confidence in a government but also in legislative terms, would have yielded a more efficient and effective government.

The second key reform behind the referendum concerned reducing the Senate’s legislative powers, modifying the (s)election of senators’, and changing the composition of the Senate. Abolishing the senate was never an option, and keeping it in some form permitted the government to maintain regional representation and interests. As noted above, reducing the Senate’s power however was argued to be fundamental to streamlining the Italian legislative process. That said, the Senate would still be consulted on matters and laws concerning constitutional reform, the electoral system, the ratification of international treaties, local and regional government, and the most important EU policies. This was an opposite approach to that of the constitutional reform approved in 2001.

The referendum also included changes to the selection of Italian senators and the composition of the senate. The Senate currently has 315 members, elected by direct popular vote, plus a few life senators (former Presidents of the Republic, and personalities appointed by the President). Under the proposed reform, the Senate would have been composed of 100 members – 95 elected members and 5 chosen by the President of the Republic, and appointed for a seven-year term (the life senators were abolished). Senators were to have been elected based on the share of Italy’s population among the 20 administrative regions: 74 of them would have been members of the regions’ legislative councils and 21 chosen among mayors. One of the most controversial and debated topics over the long electoral campaign was related to the fact that those 95 would have been elected by each region’s council «in conformity with the choices made by the voters’ and ‘in accordance with the votes and composition of each council». In other words, senators were not to be directly elected by voters but selected by regional councils. Since the referendum was rejected, the Senate will maintain its historical functions and composition.

In addition to the key reforms described above, it is worth noting some other amendments included in the referendum. In particular, the reform established that if a referendum was proposed by at least 800,000 citizens, the election will be considered valid if more than half of the total number of voters who took part in the most recent general election voted in it. This change would have reduced the threshold for referendums, as without the reform the threshold is equal to the absolute majority of eligible voters. Moreover, in terms of “direct democracy” tools, the reform stated that it would be mandatory for parliament to discuss a legislative initiative supported by at least 150,000 Italian voters (currently a popular bill proposal must be advanced by at least 50,000 voters but without any provision on the parliament’s duty to discuss it (art. 71.2 of the Constitution).

 Political and institutional consequences

Soon after the exit polls projected that the majority of Italians rejected the proposed referendum, Prime Minister Matteo Renzi quickly announced his resignation. Renzi’s resignation was indeed unusual but expected because he tied the referendum’s success to his own personal political success and support throughout the campaign. Consequently, Renzi’s risky political choice not only undermined his political career but also created another government crisis. In the week following Renzi’s resignation, the new government of Paolo Gentiloni, former Minister of Foreign Affairs in Renzi’s government, took the office. The new «president of the council of the ministers» (as such it is defined the Italian head of the government, art. 92 of the Constitution) has been appointed by the President of the Republic and he should receive the vote of confidence of… both chambers.

Renzi’s resignation also exposed the fissures and divides within the Italian Democratic Party.  Furthermore, regardless of when the next general elections will be held, the current electoral law poses several challenges. In fact, the new law approved in 2005 did not modified the Senate, as the constitutional reform modified the bicameralism and the Senate powers and relationship with the government, as said above. Therefore, the Senate electoral law was not changed also because the reform’s supporters thought the referendum would pass. Moreover, currently we will have separate units for bonus calculation (yes for the Chamber not for the Senate). Therefore, similar majorities cannot be assured.

Then, the constitutional reform was in somehow related to the new electoral law, which came into force in July 2016, albeit never been used. As for the previous 2005 electoral law, the so-called Italicum – as labelled by the Prime Minister Renzi – it is a bonus-adjusted proportional representation system. The majority bonus should be allocated to the most voted list, and no longer to a coalition of parties as in 2005. 340 MPs out of 618 are allocated to the most voted list, provided that it reaches 40% of the valid votes at national level (no further bonus is awarded if the list already had that quota through proportional distribution). If no list would get this many votes, a run-off is held two weeks later between the two most voted lists. No formal alliances (the so-called apparentamento) are allowed between lists running in the first round to compete to the run-off. Moreover, in spite of what happened in 2005, the Italicum foresaw only one legal threshold to enable access to the distribution of seats. Such access is allowed  solely for those lists that will reach at least the 3% of valid votes nationwide. Once the majority bonus is assigned (in this sense the system is majority assuring, whatever the result of the first round), the rest of seats are allocated with PR (Hare quota and largest remainders) to the list that has overcome the national threshold; no mechanism of repêchage has been introduced. Vice versa, an peculiar element of the 1948-1993 system has been re-introduced, given that the new electoral law allow voters to cast up to two preference votes (male/female candidates) for the open candidates (i.e. excluding the head-of-list) of their party, by writing the corresponding names on the ballot.

A first clarification (if any) should come from the Constitutional court, whose decision is scheduled for January 24 2017. The Court could likely drop the majority bonus for the Chamber of Deputies in order to make “more similar” the two electoral systems for the two Houses, albeit the Senate would still have a regional allocation of seats. Moreover, some changes could intervene in the voters’ provisions such as the preference votes. By the way, I am wondering if anybody among politicians and/or scholars is seriously convinced, and arguably convincing, that that PV gives more effective power to the voters in selecting their MPs.

The currently situation has changed the parties’ strategy. The Five Star Movement that firmly opposed the Italicum has quickly changed its mind by calling early elections and a vote with the Italicum electoral law. The political and social contexts offer in fact to the M5s the unique chance to probably win the run-off either against the centre-right, or against the centre-left. Vice versa part of the outgoing Democratic Party’s MPs do not dislike to have a CLPR, with a high district magnitude. However, a possible rebirth of the 1993-2005 electoral law style could give to Italian voters the chance to select MPs via SMD plus some percentage of deputies elected via CLPR. In this context, Renzi has decided to leave the Government also to avoid to be exposed to the opposition’s attacks while preparing a new electoral law. Leaving the floor to his former foreign affairs minister, Renzi – as outgoing party’s secretary – is free to prepare the campaign to obtain the new political investiture to run in the following elections.

The 2016 referendum has generated both political and institutional consequences. Another (!) electoral law could be approved in the following months (even beyond the punctual changes the Court would likely make). Under the Italian electoral sky, it seems that many things happen and nothing change. Theoretically, if the parliament would not approve any other change, it could also be possible to have general elections with two different systems for the Chamber of the Deputies and the Senate/Chamber/senate. It remains that Italy’s has not a coherent electoral law, still has two chambers with same powers, and the fact that the parliament would approve a new electoral law is not granted. The uncertainty is still there. We will (fortunately?) know more soon.

Gianluca Passarelli – Sapienza University, Roma
@gia_passarelli