The general election of March 2018 and the appointment of the so-called Yellow-Green government brought about a season of financial turmoil as well as harsh confrontation between Italy and the European institutions. The governmental coalition – composed of the anti-establishment Five Star Movement and the far-right League – pushed forward a mix of policies that raised more than some concern about the stability of Italian public finances. This lack of trust was also motivated by the belligerent rhetoric adopted by the two parties when it came to budget stability and the relationship between Rome and Brussels. Both Mr. Salvini’s League and Mr. Di Maio’s Five Star Movement repeatedly depicted financial markets as ‘a bunch of profiteers’ and the EU institutions as made up of ‘unelected technocrats’. The two leaders made it clear that the well-being of Italians and the enactment of the ‘Government Contract’ had to be put before compliance with EU economic parameters or the reactions of financial markets.
In spite of the rise in the cost of debt and the widening spread between Italian and German government bonds, the cabinet led by Giuseppe Conte kept advocating for expansionary measures, claiming that its policies would reignite the stagnant Italian economy. In its earliest months, the Conte cabinet translated words into reality and launched its two flagship measures: ‘Quota 100’, a reduction of the retirement age requirements strongly advocated by Salvini’s League and the Five Star’s ‘Reddito di cittadinanza’ (‘citizen income’), a social policy halfway between a poverty reduction measure and an unemployment benefit.
Many observers – including the EU Commission – raised doubts about the economic plans presented by Rome. The intention of the government was that the new social measures would boost participation in the job market and ultimately economic growth, thus reducing the impact of higher spending on public finances. This mantra of reconciliation between the social policies of the Yellow-Green cabinet and financial stability was, however, short-lived. The updated Economic and Financial Document (‘NADEF’), presented by the government in September 2018, foresaw much less optimistic growth, paving the way to an escalation in the conflict with the EU Commission, which was charged with overseeing national governments’ commitment to the sustainability of public finances.
For the following three months, a diplomatic marathon was run between Rome and Brussels, which ultimately led to a substantial mitigation of expenditures in the Italian budgetary law approved during the last days of December 2018. In the meantime, however, the government narrative kept claiming that the budget responded to the needs of Italians and depicted the EU rules that limit member states’ deficit and debt levels as undermining rights and increasing inequalities.
EU stability versus Italian social policy
The political conflict between the Italian government and EU institutions echoed vastly both on traditional and new media. At the core of the dispute – at least according to the narrative of the Italian government – lay an unresolvable tension between the stability of public finances (embodied by EU institutions) and the pursuit of social policies.
To shed some light on how public opinion interpreted this long-lasting, dynamic confrontation, we looked at Twitter data. This platform offers precious insights on the development of public discourse on political topics: at the cost of a certain degree of white noise, it guarantees access to unmediated citizens’ opinions as they are expressed in their tweets. To extract relevant tweets from the mare magnum of opinions expressed on a daily basis on the social network, we used a set of keywords to capture conversations about the conflict over the Italian budget.
We collected all Twitter conversation on this topic from September 27th to December 4th 2018, extracting a pool of more than 1.5 million tweets. We then built and applied thematic dictionaries – i.e., lists of specific keywords – in order to pinpoint tweets referring to the European Union (445,906), on the one hand, and tweets referring to the social policies and welfare state on the other (222,032).
Figure 1 – Twitter debate about the Italian Budgetary Law, thematic dictionaries
We focused on the intersection of the two sub-samples, that is, those tweets referring to European actors and institutions as well as to social policies (36,386 messages).
Figure 2 – Twitter debate about the Italian Budgetary Law, tweets related to EU and social Policy
As Figures 1 and 2 show, only a small portion of the online conversations on the Italian budget did focus on the relationship between the EU and social policies as such: out of 1,737,946 tweets, only 2.1% contain both EU- and welfare-related words. At first glance, this signal is ambiguous: on the one hand, the EU is not brought into play over a topic which does not fall under its mandate; on the other hand, it can also be that the relationship between financial constraints (at the EU level) and national welfare policies has not been picked up by the public.
The first step towards a clearer understanding of the content of the conversations can be made by looking at word frequencies. In Figure 3, we plot the words that appear more frequently in the reduced sample of tweets at the core of our analysis.
Figure 4 – Thematic word cloud: most recurrent words in the tweet debate
The most widely used words are rather generic in nature (spread, EU, deficit, government) as they broadly identify the topic of the conversation, but at the same time, they highlight the salience of the conflict between the Italian government and the EU institutions. These words are followed by terms that identify the policy proposals (the pension reform, ‘Quota 100’, and the citizens’ income, ‘reddito di cittadinanza’), while broader categories and concepts are relegated to the bottom of the list (rights, work, poor, health).
Figure 4 – Thematic word cloud: most recurrent hashtags in the tweet debate
The plot of the most frequently used hashtags (Figure 4) helps highlight the overall coordinates of the debate. Citizens appear to be focused on the two core policies: #RedditoDiCittadinanza and #Quota100. For the most part, the content of the conversations seems to revolve around the processual aspect of the debate, its main actors (#DiMaio, #Salvini, #Moscovici) and tropes (#DEF, #RedditoDiCittadinanza, #PublicDebt, #Quota100, #ManovraDelPopolo). Again, more substantive topics (#Pensions, #Unemployment) remain in the background.
We then looked at how frequently the words of our ‘social policies’ dictionary occurred in the same sample.
Figure 5 – Thematic word cloud: tweets about “welfare”
As shown in Figure 5, even if we consider only tweets that discuss social policies, the main concern seems to be the discussion of the new measures proposed by the government (‘Quota 100’, ‘Fornero’, ‘reddito di cittadinanza’).These references are followed by some broad terms (rights, employment, poverty), while words related to more specific policy areas (health, retirement, healthcare, school, instruction) are few.
This preliminary analysis portrays a debate in which for the most part the political news is not thoroughly debated, and the relationships between political actors, their actions and consequences are not made explicit. In order to bring to the surface how the public interpreted the political conflict between the Italian government and EU institutions, we used machine-learning methods to interpret the content of the tweets and to disentangle messages expressing a negative stance towards the EU.
Tweet debate: sentiment analysis
We used Natural Language Processing-based algorithms to classify the text of the tweets on the basis of the sentiments expressed by citizens discussing the national budget in relation to social policies and the EU. In particular, we built an algorithm aimed at identifying those messages which expressed criticism of EU actors and institutions.
Figure 3 – Twitter debate about Italian Budgetary Law and the EU, sentiment analysis
As Figure 3 shows, slightly more than 50% of the tweets in our sample express disapproval of the EU on issues related to welfare – a majority, considering that the remaining half contains both positive opinions and neutral tweets.
If we now focus only on this subset of ‘negative’ sentiment tweets, we get a different picture from the one shown before.
Plotting the keywords of our ‘social policies’ dictionary reveals a stronger focus on pensions, poverty, and social aspects in general, while the flagship policies of the Conte cabinet have relatively less importance. This suggests that criticism of the EU tends to be associated with substantive issues in the social domain, more so than messages expressing neutral or positive sentiments.
We deepened the analysis by comparing the ‘negative sentiment’ tweets with the rest of our sample by means of a machine-learning algorithm aimed at extracting the distinctive features of the messages (the results are plotted in Figure 7 below).
In this set of distinguishing words which are not necessarily the most frequently used, we can see some of the traits that are specific to the ‘populist’ narrative directed against the European institutions and financial markets. What characterizes the set of ‘negative’ messages is the reference to an elite (of a bureaucratic or financial nature) whose actions go against the interests of the people (‘sold out’, ‘slavers’, ‘victimised’, ‘profiteer’…). There is no precise reference to specific areas of welfare as the main focus of the criticisms. Rather, negative messages tend to point to the whole of the EU architecture and EU institutions, indicating that at least in some part, citizens embraced the governmental narrative of blaming the EU for imposing limits on social expenditures in the name of EU rules on fiscal discipline. This resonates with the fact that no individual actor or institution clearly emerges as the main target of the negative messages: the protests seem to point towards the Brussels’ Moloch as a whole, as well as towards a few front-page personalities (Juncker, Moscovici, Draghi).
Social media data offer precious insight into how public opinion interprets the longstanding debate on European economic governance. When discussing the 2018 budgetary law, a vast portion of the Twitter conversation we analysed expressed disapproval of the role of the EU in shaping national budget choices, in particular with regards to Italian social policies. Moreover, we brought to surface a tendency to identify European institutions and leaders with a technocratic élite whose interests are not aligned with those of “the people”.
This finding, however, has to be put in the context of a political quarrel between the EU Commission and the Italian government, the narrative of which – at least in the domestic arena – has been actively influenced by the latter.
After a summer of political turmoil, leading to the crisis of the “Yellow-Green” majority, in September 2019 Giuseppe Conte has been charged to form a new cabinet (the Conte II, as the press had it) made up of the centre-left Democratic Party (PD) and the 5Star Movement.
The appointment of a new executive marked the beginning of a “honeymoon” phase in the relationship between Italy and EU institutions. The jury is still out on whether this positive climate translates into public opinion.