Voters rally around the incumbent in the aftermath of terrorist attacks: evidence from multiple unexpected events during surveys
Voters rally around the incumbent in the aftermath of terrorist attacks: evidence from multiple unexpected events during surveys (DOI: https://doi.org/10.1017/psrm.2025.10021) was written by Albert Falcó-Gimeno, Jordi Muñoz, and Roberto Pannico in 2025. It was published in Political Science Research and Methods.
The authors aim to estimate the rally around the flag effect for incumbents in Spain after ETA terrorist attacks.
The authors contribute two quantitative analyses.
The first builds upon the UESD framework. They identify a large set of events which are expected to have equivalent effects. By estimating the pooled effect, they expect to average out the confounding effects if collateral events. Chiefly, they expect to average out the confounding effects of how terrorist attacks are reported in the media, and how authorities respond to the attacks. They use fixed effects for each event.
Rather than assuming uniform treatment, the authors model the interaction of treatment and region. (That is, whether the respondent's region matches the affected region.) The assumption being that salience of a terrorist attack differs based on proximity. The authors also use fixed effects for each region.
The authors find support for the incumbent rises 3%-5% after a terrorist attack in the same region.
The authors also test for an effect on opposition support, and find that support falls 1%-4%. This is interpretted as: the rise in support for an incumbent after a terrorist attack is partly a movement of support from opposition to incumbent, but also partly an activation of nonpartisans.
The authors also test for an effect on political identification, exploring the possibility of a conservative shift in ideology after a terrorist attack. There is no statistically significant effect.
Lastly, the authors split the analysis by the party of the incumbent: specifically between PSOE and PP. The authors find that the rally around the flag effect is consistent for conservative incumbents but not statistically significant for liberal incumbents.
The second analysis is a survey experiment. The key metric is a Likert scale: "In difficult times it is important that we all support the incumbent government, regardless of its political leaning, to face the threats that may affect the country." Some respondents are treated with a vignette about ETA.
The survey was administered using the Netquest web panel. The sample was forced to be balanced by quotas on age and education. There are 1200 respondents total.
The treatment has a strong effect on agreement with supporting the incumbent.