Do Americans Use Guns Defensively 2.5 Million Times a Year? What the Data Shows

In American political discourse, few statistics are cited as frequently by gun rights proponents as the claim that law-abiding citizens use firearms defensively up to 2.5 million times per year. Cited in congressional hearings, court filings, media commentaries, and policy briefs, this figure serves as a central pillar of the argument that widespread private gun ownership serves as a net positive for public safety, deterring criminals and protecting vulnerable citizens ^{[1], [5]}.

The figure originates primarily from a influential 1995 study by criminologists Gary Kleck and Marc Gertz, who conducted a nationwide telephone survey estimating that U.S. civilians engaged in defensive gun use (DGU) between 2.1 million and 2.5 million times annually ^[1]. The debate was refreshed in recent years by Georgetown University political scientist William English, whose 2021 National Firearms Survey estimated approximately 1.6 million annual defensive uses by gun owners ^[5].

However, when criminologists, public health researchers, and statisticians cross-examine these survey figures against official crime records, hospital emergency room data, and long-running federal victimization surveys, a stark contradiction emerges. Understanding why these figures diverge by more than twenty-fold requires examining how social science measures rare events, the mechanics of survey bias, and what authoritative data actually reveals about defensive firearm use in America.

What the Data Shows: The Multi-Million Claim vs. Crime Baselines

The fundamental challenge with claims of 1.6 million to 2.5 million annual defensive gun uses is that they conflict directly with the total number of violent crimes and burglaries occurring in the United States. Empirical researchers evaluate DGU statistics by comparing survey claims against population-level crime baselines established by the Federal Bureau of Investigation (FBI) and the Bureau of Justice Statistics (BJS) ^{[3], [4]}.

According to the BJS National Crime Victimization Survey (NCVS), which interviews approximately 240,000 individuals in 150,000 households annually, nonfatal violent crime victims protect themselves with a firearm in fewer than 1% of incidents ^[4]. Across multi-year tracking, the NCVS consistently estimates between 60,000 and 100,000 defensive gun uses per year in the United States ^[4].

In a seminal critique published in the Journal of Criminal Law and Criminology, David Hemenway, director of the Harvard Injury Control Research Center, demonstrated why the 2.5 million estimate fails basic mathematical sanity checks ^[2]:

• Burglary Discrepancy: The Kleck & Gertz survey implied that victims used firearms in self-defense during approximately 840,000 burglaries annually. However, NCVS data showed that residents were home during only about 15% of burglaries (roughly 500,000 incidents total), and awake in even fewer. It is mathematically impossible for victims to use guns in 840,000 burglaries when residents are present in only half that number ^[2].
• Gunshot Injury Claims: Kleck & Gertz respondents claimed to have shot or wounded perpetrators in about 20% of DGU incidents, implying over 200,000 criminals are wounded by armed civilians annually. Yet national medical registry and emergency department datasets (such as the CDC's WISQARS) document fewer than 20,000 nonfatal firearm injury admissions per year from all assault sources combined ^{[2], [3]}.
• Homicide Ratios: Out of millions of claimed defensive encounters involving armed confrontations, official police records document only 200 to 300 justifiable homicides by civilians per year nationwide according to FBI Uniform Crime Reporting (UCR) data ^{[3], [4]}.

Methodological Breakdown: The "False Positive" Problem in Rare Event Surveys

Why do private telephone and online surveys produce numbers that are twenty to thirty times higher than official crime statistics? Criminologists and survey methodologists point to a well-known statistical phenomenon: false positive recall bias in rare event polling ^{[2], [3]}.

Defensive gun use is a relatively rare event in the general population. In a typical national survey of 2,000 adults, if just 1% of respondents claim to have used a gun defensively over the past year (20 people), extrapolating that 1% across 260 million U.S. adults yields 2.6 million events ^[2].

Because the true prevalence is low, any small asymmetry between false positives and false negatives corrupts the total estimate exponentially:

As Hemenway noted, "If 1% of respondents misreport a rare event (false positives), and 0% of true non-events misreport, the survey will estimate a 1% rate even if the true rate is 0.01%—overestimating reality by a factor of one hundred." ^[2]

The Full Picture: Underreporting Concerns and Valid Counterpoints

While the claim of millions of annual DGUs is statistically flawed, fair-minded analysis must acknowledge that the federal government's lower baseline (~60,000 to 100,000) also faces genuine measurement challenges ^{[3], [4]}.

Proponents of higher estimates and neutral criminologists highlight several reasons why the National Crime Victimization Survey (NCVS) may undercount legitimate defensive gun uses ^{[1], [3], [5]}:

• Government Interview Reluctance: The NCVS is administered by the U.S. Census Bureau on behalf of the Department of Justice. Citizens who carrying firearms without a permit, brandished a weapon in a strict jurisdiction, or possess unregistered firearms may be unwilling to admit firearm involvement to federal government interviewers ^{[1], [3]}.
• Screening Sequence Restrictions: In the NCVS, respondents are only asked about self-defense actions *after* they report being the victim of a completed or attempted crime. If a citizen displays a firearm and deters a prospective offender before a crime is formally initiated, the respondent may not classify the encounter as a "victimization" and will never reach the self-defense question sequence ^[3].
• Deterrence Without Firing: In many genuine defensive situations, a firearm is simply displayed or mentioned without being fired. Both Kleck & Gertz and English emphasize that brandishing a weapon resolves the vast majority of defensive situations without injury, making police documentation rare ^{[1], [5]}.

In a comprehensive review conducted by the National Academies of Sciences, Engineering, and Medicine (2013), researchers synthesized available literature and concluded that while private surveys overestimate DGU frequency and NCVS likely undercounts it, defensive gun use is nonetheless a real phenomenon that occurs tens of thousands to hundreds of thousands of times annually ^[3].

Similarly, the RAND Corporation’s multi-year Science of Gun Policy initiative evaluated decades of empirical research and determined that the evidence regarding the net causal effect of defensive gun use on overall violent crime or injury reduction remains inconclusive ^[6]. RAND noted that while individual armed citizens successfully protect themselves in specific encounters, household firearm availability simultaneously increases the risks of accidental shootings, domestic homicides, and firearm suicides ^[6].

Conclusion

The conservative talking point that Americans use firearms defensively 1.6 million to 2.5 million times per year is fundamentally an artifact of flawed survey methodology. By relying on unverified self-reported phone and web surveys and multiplying tiny response fractions across the national population, these high estimates produce figures that are mathematically incompatible with known totals for burglaries, robberies, gunshot injuries, and justifiable homicides ^{[2], [3]}.

At the same time, federal victimization data indicating ~70,000 annual defensive uses represents a conservative lower bound, as some citizens underreport firearm involvement to government interviewers or experience successful crime deterrence prior to crime completion ^{[3], [4]}.

Empirical evidence suggests that defensive gun use in America occurs in the range of tens of thousands to roughly one hundred thousand incidents annually. While firearms undoubtedly play a protective role for thousands of victims each year, claims that defensive gun uses outnumber violent crimes by millions do not withstand rigorous data scrutiny.