EXPECTATIONS AND SALIENCE OF IMMIGRATION: Evidence from Italy on the political effects

In many European countries, including Italy, the estimated presence of immigrants is well above that portrayed by real data. In addition, while official statistics show that the actual number of migrants landed in the ports of Italy, Spain and Greece declined in the last few years, immigration continued to populate the political debate and to influence public opinion.

New research by Davide Bellucci, Pierluigi Conzo and Roberto Zotti, to be presented at the annual congress of the European Economic Association in August 2020, argues that voting preferences are not only shaped by the local share of immigrants, but rather by the expectations about refugees’ inflows, as boosted by the news covering arrival episodes and the public debate on the issue. 

This study is the first to examine how political outcomes are influenced by the expectations and salience of immigration. So far, previous research has documented significant effects of real immigration on electoral outcomes. This analysis diverges from the previous work, focusing on the expected presence of immigrants rather than real immigration. More specifically, the study analyses how Italian voters’ behaviour is affected by pre-electoral sea arrivals of refugees.

On arrival, refugees cannot freely go to the desired municipality, and therefore they can neither interact with natives nor influence the socio-economic dynamics in the voting municipality. But arrival episodes gained importance in the media especially before the elections, thereby increasing salience of immigration in the upcoming political competition, and influencing voters’ expectations about future arrivals. Furthermore, landing episodes were often used by far-right parties to represent immigration as a threat for natives, aiming at influencing voters’ attitudes and/or political preferences.

Using electoral outcomes of the Italian and detailed information on immigrants arrived by boat at Italian ports, the study documents that, in municipalities where refugees are more expected to arrive, voting participation decreases by 0.55 percentage points, protest votes increase by 0.31 percentage points and support for anti-immigration parties, populist parties and Northern League increases by 0.55, 1.19 and 0.86 percentage points, respectively.

These are non-negligible effects considering that, on average, turnout is around 65%, the share of protest votes is around 3.8% and the anti-immigration parties’, populist parties’ and Northern League overall vote shares is around 4.3%, 6.3% and 4.1%, respectively.

Furthermore, the study shows that immigration salience – measured as the frequency of immigration-related tweets – raises sharply in proximity of elections, especially in areas where expectations of new refugee arrivals are higher, and does not mirror the real trend of arrivals.

Since the aforementioned political effects of arrivals are larger where salience of immigration is high, the upsurge of populist and anti-immigration parties can be most likely explained by voters perceiving future arrivals as a key policy issue. The misalignment between these perceptions and actual immigration trends, however, suggest that voters’ expectations are indeed sensitive to the local electoral cycle. 

Overall, the findings of the study suggest that, as immigration becomes central in electoral disputes, expected immigration, jointly with increased insecurity and socio-economic costs of hosting refugees, arise. Representation of immigration as a permanent crisis in the media might have influenced beliefs about new inflows of immigrants, and hence boosted political support towards anti-immigration parties, while enhancing disappointment towards mainstream parties. 

Hence, strong anti-immigration campaigns might have been successful for far-right, populist parties, which, by emphasising the severity of the arrivals and proposing restrictive policies to solve the alleged refugee crisis, gathered a larger share of votes in the cities where refugees were more expected to arrive.


Authors: Davide Bellucci1, Pierluigi Conzo1,2 and Roberto Zotti1

[1]Department of Economics and Statistics ‘S. Cognetti de Martiis’, University of Turin – Campus Luigi Einaudi, Lungo Dora Siena 100A, Turin, Italy.

[2]Collegio Carlo Alberto, Piazza Arbarello 8, Turin, Italy.


Roberto Zotti
Assistant Professor of Public Economics (tenure track)

Department of Economics and Statistics ‘S. Cognetti de Martiis’, University of Turin – Campus Luigi Einaudi, Lungo Dora Siena 100A

University page: http://www.est.unito.it/do/docenti.pl/Alias?roberto.zotti#profilo

Personal page: https://sites.google.com/view/roberto-zotti