Much of the public disagreement about whether political violence is rising and who is responsible for it stems not from ideology alone but from competing definitions about what should be counted.
In scholarly literature, political violence is generally defined as the use or credible threat of physical force for political ends. Those ends may include coercion, intimidation, punishment, or symbolic signaling intended to influence public behavior, political outcomes, or institutional authority. Under this definition fall assassination attempts, armed assaults on public officials, bombings, arson attacks, violent intimidation, riots against buildings with explicit political aims, and organized plots to use force against political targets.
Just as important are the exclusions. Political violence does not include ordinary criminal activity lacking political motivation, interpersonal violence driven by private disputes, or lawful protest—even when protests are disruptive, angry, or unlawful. Nor does it include harsh rhetoric alone, however inflammatory, unless it is paired with a credible threat of force. These boundaries matter, because collapsing political violence into crime or protest obscures its distinct causes and consequences.
The hardest problem is determining intent. Political violence is defined not only by what happens but by why it happens—and intent is often ambiguous. When a crowd burns a police precinct, is the act political rebellion, criminal opportunism, or both? When a gunman attacks a synagogue, is the motive ideological antisemitism, personal grievance, or mental illness layered atop political belief? When an individual sends a death threat to a lawmaker, is it expressive rage, partisan trolling, or preparation for real-world violence?
Reasonable scholars answer these questions differently, often using dissimilar evidentiary thresholds. Some datasets require explicit ideological statements; others infer political intent from targets, symbols, or timing. Some classify riots as political violence only when their organizers articulate political goals; others focus on behavior rather than rhetoric. None of these choices is neutral, and each shapes the resulting picture.
These definitional choices explain why prominent datasets sometimes appear to disagree. Federal law enforcement datasets emphasize lethality, capability, and threats to national institutions. Protest-oriented datasets capture frequency and geographic spread. Terrorism databases prioritize ideological framing and tactical similarity to global movements. Each captures something real—and each misses something important.
For this reason, no single dataset can settle the debate. Accordingly, my work triangulates across several sources, combining federal threat assessments, incident-level databases, congressional security data, and survey research. The goal is not to produce a definitive body count but to identify patterns that recur across independent measures: who is targeted, how violence is carried out, how lethal it is, and how those patterns have changed since 2016.
This approach does not eliminate uncertainty. But it does attempt to move beyond partisan narratives and toward a more accurate understanding of how political violence in America has evolved—and why arguments over its meaning have become almost as intense as the violence itself.
The most authoritative window into political violence in the United States comes not from media tallies or advocacy groups but from the federal agencies charged with tracking national security threats—above all, the Federal Bureau of Investigation (FBI) and the Department of Homeland Security (DHS). Yet even here, readers encounter a fundamental problem: There is no single, formal US government definition of “political violence.” Instead, federal agencies operate with overlapping categories—domestic terrorism, domestic violent extremism, hate crimes, threats against officials, and antigovernment violence—each shaped by legal constraints, investigative mandates, and statutory language rather than a unified conceptual framework.