Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
Fact check: What are the most notable instances of left-wing and right-wing violence in the US since 2020?
Executive Summary
The analyses point to a contested shift in the landscape of U.S. political violence: a CSIS study reports an increase in left-wing attacks and plots outnumbering far-right incidents in the first half of 2025, while critics say the numbers are small and methodological issues make the trend uncertain [1] [2]. Independent commentary links the phenomenon to broader political polarization and elevated risks around elections, recommending interventions across policing, electoral integrity, and prevention [3].
1. A startling headline — CSIS says left-wing plots surged in early 2025
The CSIS study, published September 25, 2025, asserts a notable shift: left-wing terrorism activity rose and, in raw counts, surpassed far-right attacks and plots during the first half of 2025, though the paper emphasizes that left-wing incidents remained generally less lethal than right-wing attacks and that the decline in right-wing activity was especially pronounced [1]. This claim reframes the dominant post-2020 narrative that focused primarily on far-right threats and carries implications for analysts, law enforcement, and policymakers determining resource allocation. The CSIS finding is significant because it relies on a dataset covering recent incidents and intends to quantify directional changes rather than assert parity in lethality or organizational capacity, but the study itself admits the overall lethality of left-wing attacks remains low, a critical qualifier for interpreting operational threat levels [1].
2. Pushback and the limits of small-number trends — NPR’s cautionary read
A subsequent NPR report, dated October 25, 2025, challenges the CSIS conclusions and warns that small absolute numbers and methodological choices make firm conclusions premature, arguing that the interpretation of the CSIS data as signaling a new, comparable threat from the left is contested by experts [2]. NPR highlights criticisms centered on classification rules, selection bias, and the statistical instability that arises when analysts draw trends from limited incident counts, noting that partisan or media incentives can magnify the perception of a change that may be noise rather than signal [2]. This critique urges caution: policymakers who reallocate counterterrorism resources based on volatile short-term counts risk missing persistent, high-lethality threats even as they respond to emergent patterns.
3. Contextualizing political violence — structural drivers identified by experts
Rachel Kleinfeld’s January 17, 2024 essay situates the empirical debate within longstanding structural drivers, arguing political violence in the U.S. has been elevated by polarization, identity-based divisions, and social-psychological factors that raise the risk of election-related violence and other forms of political conflict [3]. Kleinfeld recommends a suite of interventions—improving election credibility, refining electoral rules, enhancing policing practices, investing in prevention, and tempering political rhetoric—to reduce the underlying risk, emphasizing that counting incidents is only one piece of understanding contagion, recruitment, and escalation dynamics [3]. Her analysis frames the CSIS and NPR findings as complementary rather than mutually exclusive: empirical counts may show short-term variance while structural risk factors explain why flare-ups recur.
4. Where data and inference diverge — methodological trade-offs matter
The juxtaposition of CSIS’s quantitative claim and NPR’s methodological critique reveals a core analytic tension: incident counting versus interpretation of significance, with CSIS presenting directional data for early 2025 and NPR warning about overreading sparse events [1] [2]. The debate underscores key trade-offs: relying on raw counts captures emergent activity but can mislead if sample sizes are small; qualitative threat assessment accounts for lethality and intent but can undercount diffuse or decentralized actors. Analysts must reconcile these approaches by combining incident datasets with contextual indicators—target selection, attack sophistication, and organizational links—to determine whether documented incidents represent transient spikes, strategic shifts, or measurement artifacts [1] [2].
5. What decision-makers should take from the disagreement
Policymakers and the public should treat the CSIS and NPR findings as complementary inputs rather than decisive verdicts: the CSIS data signal an observable change in incident counts in early 2025 that warrants attention, while NPR’s methodological concerns and Kleinfeld’s structural analysis argue for measured, systemic responses focused on resilience and prevention [1] [2] [3]. Practical steps include refining classification methodologies, investing in longitudinal data collection to reduce statistical volatility, and pursuing the non-kinetic interventions Kleinfeld recommends to address root causes; this balanced posture prevents both complacency about evolving threats and overreaction to limited datasets.