Can left-wing violence be compared to right-wing violence in terms of frequency and severity since 2015?

1. A surprising 2025 finding that demands scrutiny — what changed?

A Center for Strategic and International Studies (CSIS) analysis published in 2025 flagged a notable rise in incidents linked to left‑wing causes and concluded that 2025 was the first year in over 30 years when left‑wing attacks outnumbered far‑right attacks, a finding the CSIS team presents as evidence of a shifting threat landscape ^[1]. That claim is striking because it contrasts sharply with longstanding datasets showing far‑right actors caused the majority of U.S. domestic terrorism deaths. The CSIS report’s increase is presented as a frequency change, not a demonstrated reversal in lethality, and the report itself, dated 2025‑09‑30, acknowledges the attacks it counts are often lower‑lethality and less organized than those attributed to right‑wing actors ^[1]. This raises immediate questions about coding, definitions, and sample selection that affect comparability across studies.

2. Multiple datasets still show right‑wing attacks are deadlier and more numerous overall

Independent analyses and aggregated incident tallies continue to show right‑wing extremists account for a disproportionate share of fatalities since 2015. A PBS analysis and other research summarize that right‑wing violence produced roughly 75–80% of domestic terrorism deaths in the United States over recent years, while left‑wing incidents represented a much smaller share of fatalities and incidents ^[2]. Just Security’s review of multiple datasets through mid‑2025 cites 13 deaths from left‑wing attacks versus 112 from right‑wing attacks and 82 from jihadist attacks in the U.S., concluding that right‑wing plots and attacks have been both more frequent and more lethal in practice ^[3]. These figures underline that a numerical uptick in low‑severity left‑wing incidents does not translate into equivalent human cost.

3. Media attention and language trends complicate perceptions of frequency

Analyses of news language and media framing show far‑right terminology has dominated coverage since 2008 and accelerated after 2015, while far‑left mentions rose more modestly and later, meaning public attention is uneven and shaped by coverage patterns rather than raw incident counts ^[4]. The MDPI study ^[5] measured media mentions, not attacks, and therefore cannot be used alone to claim comparative violence levels; it does show that what journalists write about has diverged from what some event datasets record, contributing to mismatched public perceptions about which side is more active or dangerous ^[4]. This explains why surveys find high public concern about both poles even when empirical harms are asymmetric.

4. Methodological disputes drive divergent headlines — definitions, coding, and selection matter

Scholars repeatedly warn that inconsistent definitions of “terrorism” and fragmented databases produce divergent conclusions. An Annual Reviews synthesis ^[6] and multiple commentaries highlight that far‑right homicides outnumber left‑wing homicides since 1990 by a wide margin, but also note that comparing trends since 2015 is hampered by incomplete reporting, subjective incident coding, and gaps across federal, state, and open‑source datasets ^[7]. Critics of the CSIS finding argue the report’s methods favor counting a broader set of left‑wing actions and thus inflate incident frequency relative to more conservative tallies; proponents say broader inclusion reveals emerging patterns ^{[8] [1]}. The disagreement is methodological, not merely political, and it shapes whether one reads the data as “left surge” or “difference in measurement.”

5. Public impressions diverge sharply along partisan lines and matter for policy

Survey data show Americans widely believe politically motivated violence is increasing, but partisan lenses drive which side people see as the greater threat: majorities of Republicans point to left‑wing extremism as the primary worry, while majorities of Democrats point to right‑wing extremism ^[9]. This polarization affects policy choices, resource allocation, and which datasets gain traction in public debate. Experts urge policymakers to prioritize harm‑based metrics (fatalities, high‑lethality plots, capacity) over headline counts if the goal is to reduce deaths and major attacks, since counting low‑severity incidents without context can distort threat assessments ^{[3] [7]}.

Conclusion: across multiple, recent sources, the best reading is that left‑wing violence has shown localized increases in incident counts in some 2025 analyses, but right‑wing violence remains the more lethal and consistently dominant threat in the period since 2015; methodological disagreements, media framing, and partisan perception explain much of the controversy ^{[1] [2] [3] [7] [4] [9]}.