Main takeaways: (1) Affective polarization predicts higher support for political violence at extreme levels; (2) its main effect may be expanding opportunity structures for a small subset willing to act violently. US & Brazil show highest risk.