By C.J. Riley \ Political Theorist Series

Summary:
Statistics about hate and extremism are often treated as hard facts. But what if the way we define “violence” and “extremism” changes the outcome? This article explores why data from advocacy groups and government agencies reflect bias. The bias occurs not through falsification, but by definition.

---

In a nation deeply divided, numbers have become a battlefield. Every year, reports claim to show which side of the political spectrum commits more hate crimes or extremist violence. Organizations like the Anti-Defamation League (A.D.L.) and the Southern Poverty Law Center (S.P.L.C.) are often cited as trusted sources. Their findings end up in headlines, congressional hearings, and policy debates.

But when we look closer, a key question emerges: What exactly are they counting?

The Power of a Definition

The way we define “extremism” or “hate” can dramatically change what the data show. Some reports include online harassment or vandalism; others count only physical attacks. Some link crimes to ideology based on social media posts, while others require direct evidence of motive.

That means two reputable organizations can study the same events and reach opposite conclusions. The data aren’t necessarily wrong — they’re just built on different assumptions.

As political theorist Francis Fukuyama once wrote, “When measurement becomes moral judgment, numbers turn into narratives.”

The A.D.L. has Issues

For Example, take the A.D.L’s annual “Murder and Extremism” report. Each year, it concludes that far-right representatives are responsible for most extremist killings in the United States. The numbers are widely quoted and often accepted as fact.

However, researchers have pointed out that the A.D.L.s data are privately collected and not peer-reviewed. The organization decides which incidents qualify as “extremist,” often using a broad definition. For instance, a violent offender who once shared far-right material online might be classified as “right-wing.” This classification occurs even if the actual motive for the crime was personal or ideologically far left .

Meanwhile, acts of political violence linked to far-left or anarchist groups are often categorized differently. Sometimes, they are labeled as “protest violence” or “civil unrest.” Which by definition is political.

That’s deception. It’s the consequence of a flexible definition. When advocacy and data collection share the same space, the line between analysis and activism blur.

The Broader Problem

This challenge goes beyond the A.D.L. The FBI’s hate-crime statistics depend on voluntary reporting by thousands of local police departments. Some don’t submit data at all. Others apply different criteria for what qualifies as a hate crime.

Media-based data projects rely on news reports. This includes projects by the Washington Post or Armed Conflict Location & Event Data Project (A.C.L.E.D.). And because the media itself frames events through political and cultural lenses, those biases can carry into the data.

Economist Thomas Sowell once said, “Data are not magic; they are the footprints of human vision.” When the vision changes, the footprints do too.

When Data Meets Advocacy

Advocacy groups play an important role — they bring attention to real threats. But advocacy and analysis don’t always mix cleanly. Organizations like the A.D.L. or S.P.L.C. have moral missions as well as research arms. Their job is not just to describe reality, but to improve it.

That can unintentionally influence how they collect and present information. Emphasizing one type of violence over another might highlight a real danger. However, it can also feed public perceptions that one side is uniquely violent while others are ignored.

As Sowell argued, incentives matter. For any organization, producing data that reinforces its mission helps sustain funding, influence, and credibility. That doesn’t mean numbers are fabricated — it means they reflect Fabrication for purpose as much as precision.

The Need for Transparency

If data guide national conversations, then those who produce it must meet high standards. Policymakers and citizens alike deserve clarity about:

Definitions: What counts as “extremism,” “hate,” or “political violence”?

Sources: How are incidents verified, and who does the classifying?

Context: What kinds of cases are excluded from the data?

Without that transparency, statistics can easily become political tools rather than public goods.

Political philosopher Max Weber warned that bureaucracy often replaces moral judgment with procedure. This gives an illusion of objectivity. However, bias persists underneath. That’s the danger of overtrusting numbers: they can look neutral while quietly carrying the weight of human judgment.

Conclusion

The A.D.L., S.P.L.C., FBI, and others perform valuable work in tracking hate and violence. But as with all institutions, their data deserve the same scrutiny we apply to government or media.

The issue isn’t that the numbers are fabricated. It’s that numbers are shaped by the people who define them.
Definitions need to become more transparent and consistent. Until then, statistics about political violence will continue to reflect our politics. They will also show as much about our reality from the untrustable reports.

---

Sources:
ADL “Murder and Extremism in the United States” (2023); FBI Uniform Crime Reports (2022); ACLED U.S. Event Data (2023); Thomas Sowell, Knowledge and Decisions (1980); Max Weber, Politics as a Vocation (1919); Francis Fukuyama, Political Order and Political Decay (2014).