Every year, official crime statistics are published, cited by politicians, reported by journalists, and used to justify enforcement budgets, policy positions, and public fear. And every year, those numbers systematically exclude the most economically significant crimes committed in society — not through clerical error, but through definitional choices embedded in measurement systems for over a century.

The FBI’s Uniform Crime Reports, the primary source for American crime statistics, estimate that property crimes — burglary, theft, motor vehicle theft — cost victims approximately $17 billion annually. The Economic Policy Institute estimates annual wage theft in the United States at somewhere between $50 and $60 billion. Wage theft is the practice of failing to pay workers wages they are legally owed: unpaid overtime, minimum wage violations, illegal deductions, forcing employees to work off the clock. It affects an estimated 2.4 million workers each year in the ten most populous US states alone.

Wage theft does not appear in the Uniform Crime Reports.

The Architecture of Exclusion

What gets counted as crime, and what does not, is not a neutral decision. The categories that appear in official statistics reflect a combination of historical legal classification, enforcement priorities, and implicit assumptions about whose behavior constitutes a social threat.

Violent crime and street-level property crime are measured because they are what policing was historically organized around. Policing was organized around them because they were concentrated in observable public spaces, committed by people without institutional power, and visible to the community in ways that created political pressure to act. Corporate violations, regulatory offenses, and financial fraud were later in history classified as crimes — and much later given anything approaching equivalent enforcement resources.

The result is a measurement system where a person who steals a car appears in the national crime database, and a company that systematically shaves overtime hours from 10,000 workers does not — unless federal labor enforcement agencies happen to investigate and successfully prosecute, which requires resources, legal authority, and political will that are not reliably present.

Criminology Has Known This for Decades

The limitations of official crime data are not news to researchers. Edwin Sutherland coined the term “white-collar crime” in 1939, in a presidential address to the American Sociological Society, precisely to name the gap between what the official system counted and what empirically constituted harmful behavior. Victimization surveys — which ask people directly whether they have experienced crimes, regardless of whether they reported them — consistently reveal that recorded crime is a fraction of actual crime.

Dark figure research, the study of unreported and unrecorded crime, suggests that for most categories, official statistics capture somewhere between 40 and 60 percent of actual offenses. For wage theft, sexual violence, and corporate fraud, the figure is considerably lower — because victims either don’t know they’re victims, don’t report because they fear employment consequences, or because no category exists in which to record the complaint.

The Political Economy of Measurement

Measurement is not apolitical. The decision to count some behaviors as crimes and not others, to allocate enforcement resources to some categories and not others, and to publish statistics that emphasize certain forms of harm reflects choices made by institutions operating within political and economic contexts.

Police departments are funded, in part, based on reported crime levels. Reporting violent and property crime creates a feedback loop that justifies current enforcement structures. Reporting corporate crime at equivalent rates would imply equivalent enforcement resources — prosecutors, regulatory investigators, financial forensics — that would require a different political consensus about which harms society prioritizes. None of this is a conspiracy. It is something more mundane and more durable: institutional inertia, category persistence, and the tendency of any measurement system to reinforce the assumptions that produced it.

What the Numbers Actually Tell You

Reading crime statistics as an objective measurement of social harm is a methodological error. They measure reported, recorded, prosecuted street crime with reasonable accuracy. They measure almost nothing about the economic harm caused by wage violation, corporate fraud, environmental pollution, and financial misconduct.

The person carrying a stolen handbag appears in the data. The executive who authorized the illegal dumping of industrial waste into a river may not appear at all, because the offense was not prosecuted as a crime — it was settled as a regulatory violation with a fine that cost less than the savings from non-compliance. Understanding crime requires understanding what the official category was designed to capture, and what it was never designed to see.