Fact check: Who kills more republicans or democrats?

Executive summary: The available analyses do not support a simple answer to “Who kills more — Republicans or Democrats?” Different measures yield different patterns: state-level homicide rates have been higher in Republican-voting states over decades, while city-level hotspots are often Democratic-led cities located in Republican states, and partisan affiliation of local leaders appears to matter little for policing outcomes. Recent public opinion and threat assessments show growing fear of politically motivated violence from both wings, with some studies reporting a recent uptick in left-wing attacks. The evidence is mixed and sensitive to the metric, scale, and time period chosen ^{[1] [2] [3] [4] [5]}.

1. What each source actually claims — headline extraction that frames the debate The analyses offer three headline claims: First, a 21-year tally asserts higher per-capita murder rates in Trump-voting states compared with Biden-voting states, implying worse aggregate lethal violence in Republican states over time ^[1]. Second, a 2025 Axios analysis highlights that many highest-homicide cities are Democratic-run but located in Republican states, suggesting local–state political conflict shapes outcomes ^[2]. Third, Harvard research contends that mayoral party has little measurable effect on policing and crime trends, undermining partisan causal narratives ^[3]. Public surveys and strategic studies add that politically motivated killings are rising and contested by both sides ^{[4] [5]}.

2. The long-term state-level pattern that favors one side numerically The Third Way analysis presents a structured long-term comparison showing murder rates in Trump-voting states exceeded those in Biden-voting states every year since 2000, with a reported 23% higher per-capita murder rate over a 21-year span ^[1]. That claim frames the question at the scale of state voting patterns and multi-decade trends, not individual perpetrators’ party registrations. The metric aggregates across diverse jurisdictions, demographics, and policy regimes, so the statistical association does not by itself prove a causal link between partisan governance and homicide rates ^[1].

3. City-level hotspots complicate the narrative and point to local dynamics Axios’s 2025 piece counters broad-state claims by noting that 13 of the 20 U.S. cities with highest murder rates were in Republican-run states, and many of those cities were governed by Democrats who frequently clashed with state officials over public safety tools and budgets ^[2]. This juxtaposition highlights tension between municipal policy choices and state-level politics, and implies context — urban poverty, policing arrangements, intergovernmental conflict — is critical. It demonstrates different geographic scales can yield apparently conflicting partisan impressions about who “experiences” or “produces” more lethal violence ^[2].

4. Local leadership research undermines simple partisan causation Harvard research released in January 2025 finds mayoral partisanship has limited measurable effect on police spending, staffing, diversity, or crime trends, challenging narratives that Democratic mayors are inherently softer on crime or that Republican mayors uniformly toughen policing ^[3]. This suggests structural factors, institutional constraints, and local socioeconomics may drive outcomes more than the ideological label of a mayor. The study weakens arguments that attribute differences in homicide solely to partisan control of city halls ^[3].

5. Political violence is a separate but overlapping phenomenon Public-opinion data from October 2025 show 85% of Americans believe politically motivated violence is increasing, with Republicans and Democrats similarly worried but blaming opposite extremes ^[4]. A September 2025 strategic assessment reports that left-wing linked political violence recently surpassed right-wing attacks in frequency, including targeted attacks on conservative figures, even as right-wing violence has not vanished ^[5]. These findings distinguish general homicide trends from ideologically motivated killings, which are tracked differently and have distinct drivers and implications ^{[4] [5]}.

6. How measurement choices produce different winners and losers The contrast across sources reveals that who “kills more” depends entirely on your metric: long-run state-level per-capita homicide rates favor one partisan label, city-level hotspot counts suggest concentrated urban problems in blue cities within red states, and extremist-attack tallies track politically motivated violence that may shift over short intervals. Each metric answers a different policy question — public safety across states, urban governance, or ideological terror — so singular claims of partisan culpability are misleading without specifying scale and timeframe ^{[1] [2] [5]}.

7. Potential agendas, limits, and omitted considerations in the reporting Each analysis carries possible incentives: long-term partisan comparisons can be used to score national political narratives, city-level reporting can foreground state-versus-city disputes, and research minimizing mayoral effects may defend institutional complexity. Key omitted considerations include demographic change, policing practices, socioeconomic inequality, and the role of nonpartisan factors like drug markets. The sources do not provide direct data on the party registration of perpetrators in homicide datasets, so attributing killers to party membership is not supported by the presented evidence ^{[1] [2] [3]}.

8. Bottom line: a clear question — but no simple partisan answer The available analyses converge on one firm conclusion: there is no single, unambiguous metric showing one party’s supporters kill more than the other. Long-term state homicide rates are higher in Republican-voting states; urban hotspots are concentrated in Democratic-led cities in many of those states; mayoral party explains little; and politically motivated violence trends have recently shifted. Any claim about “who kills more” must specify the metric, scale, and period — otherwise it risks conflating distinct phenomena and overstating partisan causality ^{[1] [2] [3] [4] [5]}.