Too incivil to polarize: the effects of exposure to mediatized interparty violence on affective polarization

ABSTRACT

When incivility between opposing political parties in the institutional arena escalates to the level of physical violence involving party militants, individuals who witness the shocking images through the media can be expected to react emotionally. But how do citizens who affectively identify with one of the clashing sides experience such an incident? Known mechanisms of polarization suggest that partisans would grow further distant in their interpretation of events, and consequently increase their animosity towards the out-group. A rare opportunity to study such a dynamic through a natural experiment, in fact demonstrates the opposite: after a right-wing mob in North Macedonia stormed the Parliament and attacked left-wing MPs, right-wing supporters’ hostile attitudes towards the opposing camp subsided.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 A conceptual misunderstanding may arise from the alternating use of the terms “mediated” and “mediatized”, as the former may imply a potential manipulation of information by media through partisan spin, unlike the latter. However, they refer to different phenomena; “mediated” is associated with exposure (as used in previous studies), while the violence itself is “mediatized”. The descriptions used henceforth reflect this distinction.

2 The experimental stimuli used in the original study where the data was collected were unlikely to confound the design, given that their focus was on an unrelated topic to the internal political struggle.

3 Although North Macedonia comprises a large Albanian minority, the original survey was intended to capture the attitudes of the two parties (SDSM and VMRO) towards immigration. This most likely excluded the participation of Albanians in the study, given that the two parties are considered representatives of the ethnic Macedonian majority in the informally consociational system of the country, where most Albanians support parties of Albanian ethnicity. Due to this, ethnicity was not measured. Also, the fact that the survey’s items were only in Macedonian language further decreases the prospect of Albanians taking part in the study. However, this is unlikely to yield biased results, since even if a small number of Albanians who might have taken part as Socialdemocrat supporters, their attitudes should not differ from those of Macedonians.

4 In the original experiment, this was measured before exposure to the (thematically non-related) experimental stimuli, eliminating any possibility that the design confounded the responses.

5 After conducting post-hoc power analysis, the actual observed power of the analyses to detect differences between groups was 0.81 and 0.98 at the level of α = 0.05 for the two respective tests that reached significance.

6 Due to the small sample size, the observed effect is likely underestimated, rather than overestimated, reducing the possibility of a “false positive”.