What motivates people in referendum voting? Advocates argue that they allow voters to express their sincere preferences on issues. Matthew Bergman and Gianluca Passarelli look at the Italian case. They argue that voting in referenda can be just as party-focused as other forms of elections
Referenda have been traditionally seen as ways for voters to circumvent parties and participate directly in the policy-making process. Last year, for example, the Swiss voted on regulating tobacco advertising, increasing VAT, and bans on factory farming. One would thus presume that individuals voted according to their consciences on such measures. Their personal beliefs, therefore, would predict their voting behaviour.
In previous research on referendum voting, we found this to be the case. Yet in our recent analysis of the 2020 constitutional referendum in Italy, we found instead that partisanship and favourability towards political leaders had independent effects on individuals' voting decisions. This indicates that voters are taking cues from party leaders.
In November 2020, Italian voters were given the opportunity to decrease the size of Italy's legislature by a third. This, advocates argued, would reduce the number of corrupt politicians and the cost of governance.
The origins of the 2020 referendum go back to 2016, when the then Prime Minister, Matteo Renzi of the Democratic Party, proposed several constitutional amendments to be approved by referendum. These included, among others, reducing the influence of the Senate, rebalancing powers between the national and regional governments, abolishing the National Council for Economics and Labour, and reducing the size of the legislature.
Renzi declared that he would resign if voters rejected the 2016 referendum proposals – which they then did, despite tending to favour the proposal outcomes
Voters tended to favour these outcomes. But Renzi declared that if voters rejected the proposals he would resign. Research shows that voters were more motivated by the prospect of removing Renzi than they were enticed by the reforms themselves. Italians rejected the referendum proposals.
In 2020, just one of those amendments was put to Italian voters: a reduction in the size of the legislature. The Democratic Party and the populist 5-Star Movement both supported the reform. The opposition centre-right coalition of Forza Italia, Lega, and Fratelli d'Italia, containing many populist elements of their own, did not campaign against it. The only major opposition came from smaller parties, which risked losing out on representation if the size of the legislative chambers reduced. This time, Italian voters voted in favour of the proposal.
We began our research on the 2020 referendum voting behaviour by examining explanations for the 2016 referendum. Our previous research informed many of our areas of focus. In 2016, discontent with elites and with the political system itself were both motivating factors.
We define ‘elite discontent’ as dissatisfaction with the current government. This includes concerns about the future of the country and displeasure at current policies. ‘System discontent’, meanwhile, describes broader unhappiness about how democracy works in Italy and abroad.
Previous research revealed that voters who were more discontent with elites voted against the 2016 referendum to signal displeasure with the current government. On the other hand, voters with higher levels of ‘system discontent’ used the referendum to indicate their displeasure with the ways policy is currently made in the Italian state.
This earlier research had also identified political interest, referendum knowledge, geographical location, and partisanship as influencing 2016 voting behaviour. This led us to work with ITANES, the Italian National Election Survey, to collect similar data on individuals in the run-up to the 2020 constitutional referendum, to determine the impact these factors had on voting this time around.
Using information from more than 2,000 respondents about their voting behaviour in the referendum, two parties stood out in the analysis. 5-Star Movement identifiers were 39% more likely to support the referendum. Democratic Party voters, meanwhile, were 8% more likely to support the referendum.
Voters who identified with the 5-Star Movement and the Democratic Party heavily influenced voting behaviour in 2020, though again, discontentment with the government and the system was also a factor
We also found a role for ‘elite discontent’. The difference in support between those least in favour of the government and those most in favour was 50%. Again, the constitutional referendum ballot seemed to serve as a referendum on the current government. To examine ‘system discontent’, we used a series of questions that often measure how anti-elite, or populist, an individual is. These examined, for instance, whether they believe that people, rather than politicians, should make decisions, or whether politicians talk too much. Going from the lowest to the highest on these populist measures increased one’s likelihood of supporting the referendum by 60%.
We then explored opinions on party leaders. While not everyone in our analysis was a member of a political party, many did have an opinion on political leaders.
To our surprise, opinions on political leaders showed a stronger correlation with voting behaviour than an individual’s partisanship. Those favouring either the governing party leaders or the opposition party leaders were more likely to support the referendum. Only voters with a high opinion of Matteo Renzi showed a tendency to not support the referendum.
Opinions on political leaders showed a stronger correlation with voting behaviour than an individual’s partisanship
Personalisation of political parties in Italy has increased dramatically since 1994, when the late tycoon Silvio Berlusconi won the election with his newly launched Forza Italia. Even prior to this, the centrality of political leaders was growing in importance. Examples include Bettino Craxi, 1980s leader of the Italian Socialist Party. Increasing electoral volatility is often related to voters' changing evaluation of individual politicians.
Our research suggests that the role of political leaders is not only crucial in legislative elections. It also plays a key role in referenda, in which voters' attitudes towards politicians might help us better understand voting behaviour.