2020 Referendum

Problems and Purpose

The Five Star Movement (M5S) -Democratic Party (PD) coalition government had identified the high cost of conducting politics as a key issue, and had previously campaigned on the reduction in representatives as a potential solution. The purpose of the referendum was to seek approval for the amendments, and opposition parties had called for a referendum as the government did not have the votes for the amendment to pass without one. The amendments were also seen as a way to streamline the democratic and legislative processes but were not as wide-ranging as the one proposed in the previous years.

Background History and Context

The Italian Constitution did not originally set the number of Deputies and Senators; this precedent was set in 1963, where a 2:1 ratio was imposed. However, the increase in the number of politicians at regional and European level meant that the number of parliamentarians overall was viewed as being too high. Whilst no reforms were made, the debate continued into the 21st century, when various reductions were proposed. The first such proposal to be put to a referendum was in 2016, but this failed. Bergman and Passarelli attribute this to a variety of factors which were not present in the 2020 referendum, most notably the pledge by the then Prime Minister to step down were the referendum to fail [1]. Because of a high level of discontent with what they termed to be ‘elites’, a personalised referendum campaign about how Prime Minister Renzi strongly supported the proposals, backfired and resulted in people voting against Renzi’s proposals. The referendum campaign highlighted this further. Part of the package was a change in the electoral system, which would result in the largest party getting extra seats, but legal experts criticised this as giving the Renzi government too much power [2]. This clause was not part of the package for the 2020 referendum.  However, Bergman and Passarelli highlight a distinction between dissatisfaction with the Renzi government and with the political system as a whole; those who were dissatisfied with the political system more broadly were more likely to support the proposals [3].

The Five Star Movement party included the idea for a reduction in the number of deputies in their election platform, and after the 2018 election they entered government. Since its founding, the party has campaigned on the idea of ‘zero-cost politics’, with reducing the monetary cost of politics being one of their policies [4]. The composition of the government changed midway through the term to include the Democratic Party (PD), which were the instigators of the unsuccessful 2016 referendum. Whilst there have been numerous referendums in Italy, this was only the fourth constitutional referendum. However, all of those were held in the 21st Century.

The bill to pass the amendment did pass in October 2019 – however immediately after the passage of the bill, 71 Senators filed a request for a referendum in the Supreme Court of Cassation, which was granted. Within the Italian judicial system, the civil court is broken down into the lower and higher; the Supreme Court of Cassation is a higher court. Within this court, there exists the ‘Central Office for the Referendum’ consisting of three judges who determine whether any request meets the requirements for a constitutional referendum to be held.

Organizing, Supporting, and Funding Entities

The reform bill was approved with the support of all the major political parties on 8 October, 2019, as it was effectively voted in by the public when they showed their support for M5S. The reform was also a major part of the coalition deal struck by M5S and The Democratic Party. In this case, citizens had a major effect in making this referendum possible by voting them into office and de facto for this referendum. Following the request for a referendum being granted, most political parties expressed support for the approval of the reforms.

The process of holding the referendum was paid for by the Italian Government [5]. This however did not include funding for campaigns, in keeping with Italian law. However, the fact that most committees on the ‘no’ side of the question were independent meant that there was an uneven distribution of resources. The ‘yes’ campaign could take advantage of greater organisational strength as they had the support of established political parties, whereas the ‘no’ side was fractured between a variety of campaigns organised and funded by different and disparate groups of people. Unlike the experienced parties aligned with the ‘yes’ side, these groups had little experience of running successful nationwide campaigns and did not benefit from the support of figures already in the public eye. Citizens were able to volunteer with either campaign(s) in theory, and politicians also endorsed either side.

Participant Recruitment and Selection

Participants were not formally ‘recruited’ or ‘selected’, but as with previous referendums all eligible voters aged 18 or over had the opportunity to cast their votes. This was the only stage that was open to the public. It was up to voters to inform themselves. The government and organisers made no special effort to encourage voters to turnout, and this was left up to the campaigns. Of course, in voting for the government at the previous election voters had already given the government a mandate to pursue these reforms.

The vote being delayed because of the pandemic gave extra campaigning time to the parties giving participants of the referendum more time to take in the campaign, though there is limited evidence to suggest this helped.

On the other hand, because all the large parties were in favour of a ‘yes’ vote, there was less funding on the no side tipping the referendum in favour of the yes result. There was no official funding stipulated by the government either [6]. Secondly, the actual turnout of 51.12% could display a certain level of apathy towards the referendum from the electorate, therefore, recruitment of participants could have been better for what was quite an important referendum.

Methods and Tools Used

This democratic innovation was a referendum, a common form of direct democracy. Italy has four types of referendum; this was a constitutional referendum, which can be requested when a new constitutional law is approved by the Italian Parliament, as per Article 138 of the Italian Constitution [7]. In this case the threshold of one fifth of Senators filing a request for a referendum was met. No threshold or quorum is required for such a referendum to pass but the vote is binding [8]. In a referendum, participants are typically required to choose one out of two or more options. Referendums are likely to be binding, either legally in the bill for the referendum to go ahead, or more commonly by convention as not implementing the result would be seen as an illegitimate action [9]. It is usually the relevant government/jurisdictional body that implements them – participants have a much lesser say in that final stage of the process. As the bill to implement the proposals had already passed and the referendum was supplying formal approval for this law, this can be considered to meet the definition of direct legislation according to Graham Smith [10].

Citizens did not have any say in formulating the proposals, which were drawn up by the PD-M5S Government as part of their coalition deal. The reform already passed the required vote in the Italian Parliament, but 71 Senators applied to the Court of Cassation for a referendum as a formal check on the government-led process, meeting the threshold for a referendum to be called. This is a standard procedure, allowed for within the Italian Constitution [11]. The Senators may have lost their jobs, were the reform to pass, and the reform would have passed, and the constitution would have been amended had the Senators’ request not been granted. The referendum was a simple yes or no question.

Given that the governing parties had already campaigned on a reform package in the previous election, the resulting gain in democratic legitimacy from a referendum may have been minimal as the governing parties could already argue they had a democratic mandate for these proposals.

There were few formal techniques deployed by the organisers of the referendum. This is in addition to the lack of public resources mentioned. These were dependent on local volunteers. An example was the ‘No’ campaign group ‘Coordinamento Democrazia Costituzionale’ – who set up a local ground committee in each region to organise public discussions, and coordinate resources [12]. There were occasional public meetings that did take place in person [13]. However, these discussions were partisan in nature, and it would not be correct to classify them as deliberative processes, or processes to improve knowledge and empathy of both sides.

In meeting the standards for good referendums suggested by Benjamin Barber [14], this referendum failed to provide structured discussion as there were no government-mandated opportunities for local debate. However, it did meet these standards [15] in limiting campaign spending and advertising restrictions [16].

What Went On: Process, Interaction, and Participation

There was a lack of both formal and informal discussion procedures as the referendum act didn’t provide for them. The number of informal discussion opportunities was also reduced by the Covid-19 pandemic, which took political attention off the referendum which had already been delayed. Additionally, the issue had already been discussed for many years – reducing the impact any small-scale processes would have. The referendum question was a binary yes/no choice. Given the consensus from the parties that reform was needed, the actual terms of the proposals of themselves bore less scrutiny than in a process where citizens have other options available to them [17]. One alternative poll method is the multi-choice ballot proposed by Benjamin Barber, which would have allowed citizens to express whether they support the amendment as-is, support the principle of the reform but not the text of the amendment, and encourage politicians to rewrite the text of the proposals, or oppose the terms rather than simply say yes or no to reform [18]. In this referendum, PD encouraged a vote in favour of the reforms, but called for an amendment to the electoral system, due to concerns about large cuts in the number of Senators in particular regions and unequal cuts generally – however unlike an immediate response in Barber’s proposed poll, there was no way of achieving this without starting a new process.

The local campaigns that did take place were less likely to foster discussion as the pandemic reduced people’s engagement with their community. More engagement was given to social media campaigns [19]. Because of pandemic restrictions on face-to-face contact and public gatherings, one of the only unaffected venues for discussion was social media. Analysis of tweets about the referendum showed that few people who tweeted about the referendum had an opinion on the reforms [20], and overall, people were equally likely to see tweets in favour of the proposals as tweets against the proposals. The discrepancy between the balanced discussion on social media could not be explained by this paper, but it might suggest that discussions were ineffective.

It might be the case also that the ‘No’ campaign made greater effort to facilitate public debate as they were unlikely to win and needed more of an opportunity to convince voters. Those organising these events on either side could be classified as facilitators, but as mentioned, these discussions were partisan in nature and first and foremost their role would be to convince others of their stance [21].

Turnout was 51.1% – 14 percent lower than the 2016 referendum on the same policy area. Voters could vote in the same ways as normally, though the referendum was held over two days to reassure and encourage those who would not turn out due to the pandemic. One paper noted that in areas where a concurrent election was held, the ’Yes’ side performed worse, and turnout was much higher in the referendum in those areas. The paper did not suggest a specific cause of this. The results were communicated in the same way as previous constitutional referendums [22].

Influence, Outcomes, and Effects

The outcome of the referendum demonstrated a clear public support for the proposed constitutional reforms, with the ‘yes’ vote receiving 70% of the cast votes, with a turnout of 51.1%. The direct implications for the Italian Parliament are the reduction of its size to include 400 members of the Chamber of Deputies and 200 Senators, and a maximum of five appointed senators for life. The result has been championed by the politicians in favor of the reform as demonstrative of public support for the anti-establishment, reform-minded ethos [23] being forwarded by politicians to reduce bureaucracy. It is seen as the first step towards further reform of the Parliamentary system. Reducing number of MPs is estimated to produce savings of 100million euros a year, making the Italian Parliament ‘more efficient and less bureaucratic’ [24], but critics highlight that this saving is minimal for the government compared to its 900-billion-euro yearly expenditure [25], with claims that 100 million euros is a small price to pay for the democratic benefits of the larger legislature.

Implications on democracy of the result include the reduced representation of constituents to a ratio of 1:151,000 and 1:300,000 between parliamentarians and citizens for the Chamber of Deputies and the Senate respectively [26]. Those who advocated for the reform highlighted that this will reduce bureaucracy within the Italian Parliament which will lead to a smoother democratic process. However, the OpenDemocracy news group has argued it has also made Italy ‘the least representative Parliament in Europe’ [27]. Further effects on democracy include a disproportionate impact on more rural regions of Italy and minorities in relation to representation with lowered number of MPs causing ‘difficulties in establishing direct relationship with voters’ [28].

In addition, the reduced size of the legislature reduces the number of opportunities for small parties to be elected, as parties would mathematically require more votes to win a single seat than beforehand. Parties who did well among voters non-resident in Italy were also critical as this group would mathematically be the least represented [29].

Analysis and Lessons Learned

In relation to inclusion the vote achieved a turnout of 51% and this represents a failure in relation to inclusion as a substantial proportion of the population did not directly engage. It must be said that the vote occurred during the COVID-19 pandemic which deterred engagement and furthermore the clear support for reform meant that participation was not necessarily deemed impactful by citizens. All regions within Italy took part in the vote and all voted in a majority in favour of reform. Turnout was however higher in areas which had other elections taking place concurrently, so the number of people voting might overstate the level of engagement. As the government made no efforts to encourage high turnouts this can be attributed to that flaw in the design of the referendum.

The Covid-19 pandemic caused issues regarding the degree of deliberation that could be organized around the referendum lowering the level considered judgement with respect previous referendums. The pandemic required significant public and government attention, this reduced the focus on the referendum causing lower levels of public debate and interest relating to the vote. Because of the COVID pandemic, the referendum was never the main political news story, and restrictions on campaigning meant that local, ground campaigning was limited. This reduced the number of opportunities for voters to gain knowledge of or empathise with viewpoints on either side. As mentioned, a lot of people who tweeted about the referendum did not express a preference, suggesting there was room for understanding of the issues to be improved. Pandemic related complications consequently caused reduced levels of public knowledge on the subject matter. All major parties campaigned for the yes side, and campaigning was limited. According to Arthur Lupia, the low information subsequently provided to voters increases the possibility of information shortcuts or heuristics informing decisions, increasing the likelihood of a misunderstanding of the consequences of their decision [30]. This paper argued that where respondents had little factual knowledge, they would take into account the preferences of organizations they did know and vote similarly. In considering this referendum, all major political parties supported the proposal. It could therefore be assumed that voters considered the fact that their preferred party supported the proposal and were more likely to vote accordingly.

There was also a lack of alternate mechanisms to mitigate the effects of the Pandemic – for example high-profile televised debates to replace public meetings. The terms of the referendum were slightly adapted from those of the 2016 referendum on the same topic. This slightly limits the transferred understanding between the two instances, limiting its impact on the democratic good Graham Smith defines as considered judgement [31] relating to the public decision, especially as there were limited opportunities provided to gain further knowledge. The lack of a strong ‘no’ campaign, in turn, shows a low level of popular control about option analysis for participants as the ‘no’ campaign was limited in their ability to advocate for their platform although the Italian public had previously demonstrated their support for reform and the issue had been present in public debate during prior referendums on the matter.

Referendums are a frequent practice in Italy and in repetition of the process, a transparent process has been achieved through refinement based on past referendums. Regarding the transparency of the process, internally there was high transparency with the bill to hold the referendum supported by all major parties in 2019 – supplying sufficient time to debate the intricacies of the vote in terms of question phrasing etc. Delays through COVID and the prior debates in 2016 furthered internal transparency. A transparent process was seen through the available public records, the public could see how the referendum came about and how the issue had been debated. One limitation was most deliberation and campaigning were informal procedures, reducing transparency as there is no such record for public viewing, but this is true of most referendums. A further strongly refined aspect of the 2020 Italian Constitutional Referendum was its efficiency, the Italian process is already streamlined, and its efficiency was furthered by other elections being held concurrently in certain reasons improving the individual efficiency of the process. The lack of readily available information for individuals to become informed on the issue could be a criticism of the referendum, however, the actual voting process was highly efficiently organized.

There is debate over the ability of direct democracy to incur actual political change, as scholars such as Gerber and Lupia have argued that this is dependent on whether the results are implemented. [32] Whilst in this aspect the Italian referendum is a demonstrable success with the constitutional changes set to be forwarded, from the perspective of Gerber the almost universal support from the major parties whose role is to implement the policy makes this use of direct democracy an ‘expensive way to seek policy change’ [33] rather than an effective use of direct democracy. Specifically, Gerber argues that referendums are only worthwhile to change policies that would have no chance of succeeding in the legislature – a far cry from the near-unanimous support from elected parties. The high support amongst the major parties for reform led to little public consultation over problem definition consequently limiting popular control in this aspect, with problem definition occurring through bureaucratic means.

The 2020 Italian Constitutional Referendum may be an effective supporting example for Broder’s theory that referendums can be used to subvert the interests of the many [34]. The policy proposal had received high democratic support, as the M5S and PD parties in government had campaigned on it, alongside the support of parties in the legislature that represented 80% of votes cast in the previous election. This high level of public support makes the 2020 referendum unprecedentedly unbalanced in public support for each side, respectively. The small number of Senators, with the interests of keeping their positions, were able to instigate a referendum, is evidence of Broder’s of conception of powerful minorities subverting normal political processes for special interests through referendums. The requisite number of signatures needed to instigate a referendum through the Court of Cassation was not met, but Senators could more easily mobilise [35].

Whilst the normal referendum processes undertaken for the 2020 Italian Constitutional referendum are replicable and transferrable in that it is an effectively replicated process within Italy, the example itself is not highly transferrable due to the flaws in its processes: the actual process of instigating the referendum via the Court of Cassation is unique to Italy. However, there is nothing to suggest that a similar referendum could not be replicated in other countries, as evidenced by the existence of similar mandatory referendums.