Assassination and Public Opinion Dynamics: Evidence from Japan

• In Event: Domestic Political Violence, Arrests, and Public Attitudes

Abstract

Among acts of political violence, the assassination of a political leader is considered to have a particularly significant impact on people’s attitudes or psychology. The risk of being assassinated is said to increase with repressive leaders in non-democratic countries (Iqbal and Zorn 2006). This suggests that political assassinations in democratic systems have a significant impact on people’s political and social attitudes because they are rare events. Indeed, the assassination of John F. Kennedy is still strongly remembered by American voters as an important event. Howard Schuman and his colleagues have found that those who were adolescents or young adults at the time of the JFK assassination still refer to the event as a particularly significant event in polls of the 2000s (Schuman and Rodgers 2004; Schuman and Corning 2012).

However, there is still room for argument about the impact of the assassination on public opinion. Because political assassinations are rare events, previous studies examined the influence of political assassinations on public opinion, focusing on a few specific incidents. Most previous studies examining the relationship between assassination and public opinion focused on the assassination of JFK or Israeli Prime Minister Yitzhak Rabin. However, it remains unclear how generalizable the findings of previous studies are, since they only focus on a few limited contexts. For example, previous studies have argued that violence against politicians increase support for the victim’s party, but Krakowski et al. (2022) analyzed the assassination of a Polish opposition politician in 2019 and found that, rather, support for the victim’s party decreased.

Furthermore, since assassinations happen suddenly, it is difficult to see how public opinion changed before and after the event. It is impossible to plan the timing of a public opinion poll in anticipation of an assassination, and most previous studies analyzed survey data obtained a short time after the event. There are a few studies that compare public opinion immediately before and after an incident. However, most such studies analyzed surveys conducted on a small sample size for a specific demographic (e.g., high school or college students). Krakowski et al. (2022) attempted to solve this problem using Twitter data, but they did not consider how much Twitter users reflect the entire electorate.

To overcome these limitations, we analyze public opinion in Japan before and after the assassination of former Japanese Prime Minister Shinzo Abe. Shinzo Abe, the longest-serving leader in Japan, was shot from behind while making a speech at a campaign rally on July 8, 2022, two days before the House of Councilors election. We examine the impact of the Abe assassination on Japanese public opinion using survey data from voting advice applications (VAA). We happened to get data that allowed us to measure public opinion before and after the assassination of Abe, as we were running the VAA with Asahi Shimbun, a major Japanese newspaper, during the election period. Our VAA data recorded 2188336 respondents by July 7, the day before the incident; 40667 on July 8, the day of the incident; 113514 on July 9, the day after the incident; and 213021 on July 10, the day of voting.

There may be some concern that estimates based on VAA data are different from those obtained from standard polls and are not generalizable. Thus, we first compare our VAA data through July 7 with a nationwide random sampling survey conducted in March-April 2022. We find a strong positive correlation between the estimates obtained from both methods. When partisanship and policy preferences are estimated by prefecture, age, and gender, the order of estimates across subgroups is roughly consistent between the VAA and the random sampling survey.

After confirming the validity of the VAA data, we estimate how public opinion changed after the assassination. Our results reveal three findings. First, public opinion shifted to the right or conservative after the assassination. Support for the Liberal Democratic Party, the victim’s party, increased slightly, while support for the left-opposition parties decreased. There was also an increase in support for strengthening Japan’s defense forces and for restricting personal or privacy rights in order to keep order. Second, our results suggest that the rightward shift influenced by the assassination of Abe occurred independent of pre-event partisanship or policy positions. Comparing partisanship and policy preferences across prefectures, their positions did not change before and after the incident. After Abe’s assassination, all prefectures shifted to the right to the same extent. Third, women were more affected by the assassination than men. Before the incident, women were more leftist or liberal than men in Japanese public opinion. After the incident, however, such gender differences narrowed because of women’s shift to the right more than men’s shift.

Authors

• Taka-aki Asano, Kansai University

• Shoko Omori, Hosei University

• Masaki Taniguchi, The University of Tokyo