COVID-19 risk perception and political actors: how confidence in the government or opposition affected risk perception during the Italian lockdown

ABSTRACTThis paper investigates the connections between the socio-demographic patterns of COVID-19 risk perception and the influence of confidence in the government or opposition during the first Italian lockdown. Confidence in political actors is used as a proxy for the adoption of the interpretative frames they provide. Socio-demographic characteristics are crucial in understanding the perception of risks ascribed to a specific event. However, the politicization of such events and the subsequent production of interpretative frames by political actors gives rise to further crucial factors when the event is characterized by ‘uncertainty’. This empirical analysis is based on primary data from a survey conducted in Italy during the last three weeks of closures (26 April−17 May 2020, N = 1,704) and clearly shows that the most worried respondents were women, elderly people, those living in the north of Italy, those with an upper-secondary school level of education or those with a job in the socio-cultural professions. Accounting for confidence in the government or opposition shows that respondents who have a higher degree of confidence in the opposition were less worried than those with a higher degree of confidence in the government. As politicization of the pandemic primarily divided those in favour and those opposed to the strict measures to limit infections at the expense of the health of the economy and personal freedom, the positions of respondents on these two trade-offs (economy/lockdown and freedom/lockdown) are accounted for. The analysis shows that position on these two dimensions completely reflects the association between confidence in political actors and risk perceptions – apart from the elderly, those resident in the north of Italy and socio-cultural professionals. These results provide insights into the literature assessing the socio-demographic patterns of COVID-19 risk perception and the politicization of the pandemic.KEYWORDS: COVID-19risk perceptionpolitical attitudessocio-demographic characteristics Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1. The associations between socio-demographic characteristics and risk perception are introduced in the first section.2. The government and opposition were the main political entities with decision-making and legislative powers during lockdown.3. Such frames define the risks associated with a disease and actual victims in different ways. Furthermore, victims can be ascribed to both the disease and the measures enacted (Staniland and Smith Citation2013).4. These governments opted for a ‘moralizing’ discursive strategy, feeding concerns about the virus with the aim of achieving compliance in order to minimize the number of infected people (Monaghan Citation2020). As seen in previous epidemics, such a frame competed with another focused on the minimization of the health aftermath of the disease to avoid social and economic strains.5. Mass media communication played a central role in the framing strategy of both the government and the opposition. For a more detailed discussion of its role, see Mazzoni et al. (Citation2021) and Battistelli and Galantino (Citation2020). Barrios and Hochberg (Citation2020), focusing on the United States, provide insights into the connections between the partisanship/political ideological orientations of individuals and their disposition to ascribe higher levels of credibility to specific actors and news sources, i.e. the ones matching their political orientations.6. Battistelli and Galantino (Citation2020) provide a thorough description of this opposition in the Italian context.7. In August 2019, the coalition government of the M5s and the League was dissolved: the League withdrew it support for the government and a new round of consultations led to the formation of an alternative coalition government consisting of the M5s and the Partito Democratico (Democratic Party, PD).8. For example, no mainstream political party criticized the medical and scientific community. The objects of contention were the management of the crisis and the policies taken to contain it (Bertero and Seddone Citation2021; Bobba and Hubé Citation2021).9. During the COVID-19 emergency, people’s evaluations of their governments proved to be associated with their compliance with regulations (social distancing, mask-wearing, hygienic measures). Lobera et al. (Citation2023) detect a non-negligible role of people’s pre-existing political values in affecting these evaluations, according to their ideological affinity with the government.10. ‘La Politica e la Sfida della Pandemia COVID-19’. The questionnaire is provided in the Appendix.11. The link was disseminated via social media, obtaining a self-selected sample. In order to exclude from the analyses those respondents not entitled to vote, the two questions focused on voting behaviour included the response category ‘Not entitled to vote’. The final dataset is available from the authors on request. Table A1 in the Appendix shows he frequency distribution of 2018 vote choice variable in the final sample.12. The six-monthly inquiry into the Italian electorate up to 31 December 2019 was used for the reference statistics. Source: https://dait.interno.gov.it/elezioni/rilevazione-semestrale. These reference statistics do not include further socio-demographic variables, such as age and education. Therefore, the weights that have been employed do not allow us to address sample distortions associated with over- and under-representation of specific social groups.13. The Italian version of the question is: ‘Per favore, indichi la Sua posizione rispetto alle seguenti alternative. Di fronte all’emergenza sanitaria: Si deve mantenere il più possibile il confinamento a casa anche se questo comporta un peggioramento delle condizioni economiche e un aumento della disoccupazione (1) – Si deve rendere più flessibile il confinamento a casa per riattivare quanto prima l’economia anche se questo può comportare una maggiore propagazione del virus (5)’.14. The Italian version of the question is: ‘In merito alle misure adottate per fronteggiare la pandemia in corso, lei dove collocherebbe la sua opinione? Lei ritiene che si dovrebbe: Contenere il contagio anche a costo di limitare la libertà personale (1) – Non applicare rigide misure restrittive sulla libertà personale anche a costo di estendere il contagio (5)’.15. In accordance with H1, relative confidence rate constitutes the key independent variable in the analyses, whereas the two attitudes are the independent variables employed to test H2 and H3.16. See Table A2 in the Appendix for mean, standard deviation, minimum and maximum values of the four continuous variables included in the regression models.17. Oesch’s (Citation2006) 8-class schema. The unemployed, retired, those unable to work and students are codified according to their last job or the job of one of their family members. According to this schema, socio-cultural professionals include those professionals/managerial jobs endowed with highly marketable skills and involving high levels of interpersonal interaction, e.g. medical doctors and university/school teachers.18. The variable is generated according to the date of response and is dichotomized between prior to and after the first ‘reopening’ measures.19. Lanciano et al. (Citation2020)provide an accurate schematization of the three phases set by the Italian government to cope with the COVID-19 emergency.20. For a thorough presentation of the results of Model 1, see Marchesi and De Luigi (Citation2022).21. It must be stressed that the coefficients pertaining to the area of residence decrease with respect to both M1 and M2 as far as the north east and centre are concerned. The difference in risk perception between the north west and the south and islands increases in M3 (−3%) with respect to M1 (−2%) and decreases in M3 with respect to M2 (−4%).22. With reference to Figure 1, it should be noted that younger and elderly respondents are associated with a lower mean value in the full model (M3) as compared with the first model (M1), whereas the opposite pattern is shown as concerns the 35–64 age group. For areas of residence: lower mean values of risk perception are observed in M3 than in M1 for the north west and the south and islands.23. The variable is recoded highlighting the number of respondents who voted for each of the three main parties, according to the official election results. Considering the sensitivity of the information, respondents were allowed to skip the question or to decide not to declare their vote choice in the 2018 general election. Furthermore, a few respondents were not eligible to vote in 2018 (i.e., they were under 18 in 2018).24. The survey was conducted in Italian. The authors are prepared both to translate, and to offer support in the translation of, questions and items on request.Additional informationNotes on contributorsAndrea MarchesiAndrea Marchesi holds the position of Research & Panel Management Executive at BVA Doxa. He attended the PhD Programme in Political and Social Sciences at the University of Bologa.Nicola De LuigiNicola De Luigi teaches Sociology at the University of Bologna and is currently the Delegate for International and new degree programmes. He is a member of the Scientific Committee of the PhD Programme in Political and Social Sciences at the University of Bologna.