Right wing or left wing protests more violent
Executive summary
Across multiple government, academic, and media analyses, right‑wing political violence in the United States has been more frequent and far deadlier than left‑wing violence in recent decades, even as some data show increases in left‑wing incidents from very low baselines; however, methodological differences and policing practices complicate simple comparisons [1][2][3].
1. Right‑wing violence: frequency and lethality
Longstanding datasets and expert reviews conclude that most domestic terrorism fatalities have come from actors on the right: several analyses place right‑wing attackers responsible for roughly three‑quarters of U.S. domestic terrorism deaths since 2001 and document numerous high‑profile lethal attacks such as Pittsburgh, El Paso, and Buffalo as recent examples [1][2][4]. Brookings and other policy papers report that since 9/11 far‑right extremists have killed more people in the United States than other ideological groups, noting both mass‑casualty events and a steady stream of lower‑level assaults and hate crimes [2]. Independent academic reviews and syntheses thus characterize right‑wing violence as both more frequent and more lethal in recent years [5][4].
2. Left‑wing violence: rising incidents but low lethality
Multiple recent studies and policy trackers record an uptick in left‑wing incidents since about 2016, but they emphasize that this rise comes from a very low baseline and has produced far fewer fatalities; some analyses count only a handful of deadly left‑wing attacks since 2020, with most left‑wing incidents targeting property or producing nonlethal clashes [3][4][6]. CSIS reports that left‑wing attacks remain overwhelmingly non‑lethal and that since 2020 only two or three fatalities can be attributed to left‑wing terrorism depending on case definitions, underlining a stark gap in lethality compared with right‑wing actors [3].
3. Nuance from comparative and global studies
Not all academic work reaches an identical conclusion: a multicountry, dataset‑based study led by researchers at START/UMD found that in some samples the probability an extremist attack is violent can be similar across ideologies, and in the U.S. sample right‑wing and Islamist attackers showed comparable propensities while left‑wing cases were less likely to be violent—findings that caution against overgeneralizing and highlight dataset and temporal effects [7][8][5]. These studies show ideological differences can shift with time, geography, and how “violence” and “terrorism” are defined, so any headline assertion must be anchored to specific datasets and years [5].
4. Policing, media coverage, and political narratives that shape perception
Analysts note that policing patterns and elite rhetoric distort what observers see: ACLED and other reviews found police used greater force at many left‑leaning protests in 2020, and political leaders have sometimes amplified the threat from one side for rhetorical effect, affecting both public perception and enforcement priorities [9][10]. FactCheck and other outlets caution that political actors have selectively cited data to portray violence as concentrated on the left or right, and that careful reading shows violence spans the spectrum even if mortality concentrates on the right [11][10].
5. Bottom line, caveats, and what’s missing
The balance of evidence in the provided reporting indicates right‑wing protests and extremist violence have been more frequent and considerably more lethal than left‑wing counterparts in recent U.S. history, though left‑wing incidents have risen from low levels and merit attention [1][2][3]. At the same time, methodological differences between datasets, evolving tactics, local policing choices, and the political incentives of some commentators mean that simple binaries understate complexity; several academic sources expressly call for continued, careful monitoring across ideologies rather than partisan shorthand [7][8][6]. Reporting here does not assess every dataset or the very latest incidents beyond these sources; where claims fall outside the provided reporting, that limitation is acknowledged.