Personalization of Politics Between Television and the Internet: Leader Effects in the 2013 Italian Parliamentary Election

ABSTRACT

This paper provides an empirical assessment of the relationship between media exposure and leader effects on voting through an analysis of Italy––an ideal case for the study of the personalization of politics and its connection with political communication. The results show the dominance of leader effects among voters strongly exposed to television and a somewhat differentiated impact on Internauts. By looking at leader effects across different audiences, this paper elaborates on the missing link between electoral research and political communication, and it eventually speaks to the broader question of how important media is for the outcome of contemporary democratic elections.

KEYWORDS:

• ICTs
• leader effects
• media exposure
• personalization of politics
• voting behavior

Notes

^1. So far, only one study by Gidengil (2011) tackled the issue in a comparative perspective—albeit with inconclusive results due to the acknowledged limitations within the available data.

^2. From a methodological point of view, sticking to survey data alone implies opting (along the lines of virtually all existing research on the topic) for an “attentiveness” approach. This approach carries some drawbacks as compared to a “linkage” approach, most notably, the inability to connect the actual content of media coverage to individual patterns of voting behavior. However, the “linkage” approach comes with problems too. According to Dobrzynska et al. (2003, p. 29), the major limitation of this approach refers to the fact that it “does not address the bottom line question, that is, how much effect the media do exert on voters’ final choice.” Given the foremost interest of this paper in the very same bottom-line question, I thus decided to resort to the “attentiveness” approach, yet keeping in mind its inherent limitations when it comes to interpretation and generalization of the empirical findings.

^3. Because the analysis deals with the determinants of vote choice, I will leave aside respondents who abstained as well as those who picked the ‘did not vote,’ ‘do not know,’ or ‘no answer’ response options. In turn, this lowers the number of cases included in the analysis down to N = 950.

^4. The choice to sum up all the items from the Internet activity battery into a single additive scale is supported by the results of a factor analysis (all items load heavily onto the same factor; Cronbach’s Alpha for the additive index equals .84). I further tested different versions of the index by excluding one activity at a time, each time obtaining the very same results.

^5. The use of personality traits in place of thermometer score evaluations relies on the idea set forth by Fiorina (1981) that the thermometer might also be capturing factors such as retrospective judgments, party influence, issue positions, and so on.

^6. Respondents have been asked to rate each political leader on the four traits on a 4-point scale ranging from “not at all” to “very much.” Each variable has then been rescaled: “0” stands for a negative attitude (i.e., “not at all” and “not very much”) while “1” stands for a positive attitude (i.e., “fairly much” and “very much”).

^7. Previous contributions highlighted that Grillo’s popularity, already high among Internauts, grew exponentially among television viewers throughout the 2013 campaign as a result of the progressive hybridization of the Italian media system (Barisione et al., 2014).

^8. Unfortunately, it was not possible to estimate the effect of coalition leader evaluations on single-activity-based subgroups of Internauts due to the very low number of cases.

^9. When testing the model on party rather than coalition choice, Berlusconi appeared somewhat less important to respondents heavily exposed to television.

^10. Bentivegna and Ceccarini (2013) show that the figure for political discussion during the 2013 election campaign was 66% among citizens who used the Internet to get political information compared to 45% for all others.

^11. It is worth highlighting that Italy ranks fourth in the world—and first among Western European countries—when it comes to average daily television viewing time per person (http://www.statista.com/statistics/276748/average-daily-tv-viewing-time-per-person-in-selected-countries/ [last accessed: 07/16/2017]). At the same time, Italy also ranks way below the EU27 average with regard to the frequency of Internet use (http://ec.europa.eu/eurostat/statistics-explained/index.php/File:Internet_use_and_frequency_of_use_by_individuals,_2013_(%25_of_individuals).png)

Additional information

Notes on contributors

Diego Garzia

Diego Garzia is Senior Researcher and Lecturer in the Department of Political Science, University of Lucerne.