NY Safe Inc. · Original data analysis
Defensive Gun Use Without Firing and the Case for Lawful Carry
In 2,204 of 2,627 national survey reports examined by NY Safe, the person who described using a gun defensively said they never fired. That finding helps explain what shooting totals miss. In New York, where carry approvals rose after Bruen as shootings fell, the right to lawful self-defense deserves a place in the public-safety debate.
By Peter Ticali · NY Safe Inc. · Research reviewed November 9, 2026
The short answer
New York City’s experience shows that broader access to carry licenses can coexist with falling shootings. The public-safety case for lawful ownership starts with the person who can stop a violent attack. Shooting totals capture harm that occurred. They give us little visibility into attacks interrupted before someone was hurt. In this unweighted analysis of one reported lifetime encounter per respondent, NY Safe excluded animal encounters and counted what people said they did and what they believed happened.
- 83.9% (2,204 of 2,627) said the defender did not fire.
- 65.8% (1,728 of 2,627) said the defender did not fire and believed the incident left no crime to report.
- 147 of 191 respondents who selected a public location reported both outcomes. The survey did not independently establish licensing status or whether the action was legally justified.
What these numbers measure. This is an unweighted analysis of national survey answers. “No crime remained” is the respondent’s judgment. The survey does not verify police calls, licensing, legal justification or what would have happened without the firearm. The New York City figures come from separate records.
Disclosure. Peter Ticali is NY Safe Inc.’s founder and lead instructor. NY Safe sells firearms training, including New York licensing courses. This editorial analysis has not been peer-reviewed.
Public survey file and questionnaire fields · How NY Safe counted the reports
What happened when New Yorkers gained greater access to carry?
When the Supreme Court struck down New York’s special-need requirement in 2022, then-Mayor Eric Adams warned that the decision would put New Yorkers at further risk of gun violence. He also said the city would take steps to mitigate that risk. Governor Hochul presented the state’s new restrictions as a public-safety response to Bruen. Those are specific public claims worth testing against the city’s experience. Mayor Adams’s June 2022 statement · Governor Hochul’s announcement
A November 2025 analysis of NYPD records by THE CITY and The Trace found more than 17,000 New York City concealed-carry approvals since the law changed, with over 8,000 additional applications pending as of October 1, 2025. That is a snapshot of city applications and approvals. It leaves the current number of active NYC carry licenses, the statewide total and how often license holders actually carry unresolved. New York law provides for a statewide license database. Albany can give the public a much clearer picture by publishing anonymous totals for licenses issued and active, by county and license type, along with renewals and processing times. The joint reporting and its licensing figures · New York Penal Law § 400.02 · NY Safe’s analysis of the NYC licensing backlog
The NYPD’s public incident data show 688 citywide shootings in 2025, compared with 1,562 in 2021 and 777 in 2019. That is a decline of 56.0% and 11.5%, respectively. Broader access to lawful carry and fewer shootings happened in the same city. The same numbers cannot tell us how much any one policy contributed to the decline. Albany should apply the same test to its own claims. A falling citywide total gives officials no basis to credit a specific carry restriction without evidence that separates its effect from other changes. NYC Open Data shooting incidents · NY Safe’s broader New York crime analysis
The NYPD’s November 9, 2026 release illustrates how the time window shapes a headline. Citywide shootings were down 10.4% from January through October, from 598 to 536. In October alone, they were up 4.4%, from 45 to 47. Both figures compare with the same period in 2025. A reader needs the dates and counts before either percentage can tell them anything about the trend. The 2026 numbers are preliminary and cover ten months. NYPD’s November 9 citywide release
New York also has an example of what the defensive option can mean for an individual. In 2016, Chautauqua County’s acting district attorney announced that his office would file no charges against a Fredonia homeowner who fired a properly licensed handgun. The district attorney cited New York’s premises-justification statute and a review of witness and forensic evidence. This is an illustrative case. The available source is a local newsroom’s account of the prosecutor’s decision. The decision concerned one set of facts, and the outcome without the gun is unknowable. WRFA’s published account of the district attorney’s statement
How Albany passed the CCIA in one day
The Concealed Carry Improvement Act moved through both chambers and went to the governor’s desk on a single date. The official record gives New Yorkers a precise timeline.
- June 23, 2022. The U.S. Supreme Court released New York State Rifle & Pistol Association v. Bruen and struck down New York’s special-need requirement for public carry. Supreme Court opinion
- June 24. Governor Hochul announced an extraordinary session to begin June 30, seven calendar days after the ruling. Governor’s special-session announcement
- July 1. The CCIA bill was referred, passed by both chambers under a message of necessity, delivered to the governor and signed as Chapter 371. It became law eight calendar days after the decision. Senate bill history
New York’s Constitution ordinarily requires the final text to be on legislators’ desks for at least three calendar legislative days before final passage. A governor’s message of necessity can bypass that waiting period if the final text is on the desks before the vote. On the Assembly floor, members questioned the emergency procedure when the principal provisions would take effect September 1. State leaders argued they needed to respond quickly to the ruling and put screening and training measures in place. New York Constitution, Article III, § 14 · Assembly floor debate · Governor’s signing announcement
In my view, using the emergency procedure to rush through a package restricting the exercise of a constitutional right was an abuse of the process. The state had the power to expedite the vote. It also owed New Yorkers time to read the ruling, study the final bill and ask whether the new restrictions would make them safer. The timing and the decision to bypass the ordinary review period are the basis for that judgment.
What the 2026 survey records about defensive encounters
William English’s second National Firearms Survey was fielded in July and August 2026. The public file contains responses from 51,398 U.S. adults. English’s analysis identifies 16,688 firearm owners through his internal consistency checks. The survey asked about defense of oneself, property, family or household members and told respondents to exclude military, police and security work. A follow-up asked whether the respondent fired, showed the gun, or did neither. A separate follow-up asked whether the respondent believed a crime remained to report, with a separate answer for animal threats. Survey author’s results and design summary · Harvard Dataverse, survey file and question wording
English reported a 76.3% no-shot share across the incidents he analyzed. We asked a narrower question, using one reported lifetime encounter per person and excluding animal encounters. Our 83.9% includes responses where the person said they neither fired nor showed the gun. Removing that category or including repeat respondents lowers the share. The sensitivity table shows how much. English’s published summary
51,398 public survey rows → 2,893 respondents reporting one encounter → 2,802 after completion and ownership checks → 2,627 after removing 175 animal-threat reports.
2,627 · Eligible single-encounter reports
2,204 · 83.9% · Reported no shot fired by the defender
1,728 · 65.8% · Reported no shot and believed no crime remained
All percentages use 2,627 as the denominator. The last bar reflects the respondent’s assessment.
| Defender action | No crime remained | Crime remained | Total |
|---|---|---|---|
| Showed the gun without firing | 1,165 | 328 | 1,493 |
| Neither fired nor showed the gun | 563 | 148 | 711 |
| Fired the gun | 308 | 115 | 423 |
| Total | 2,036 | 591 | 2,627 |
Of the 2,204 no-shot reports, 476 said a crime remained. The survey’s “neither” category gives a verbal reference to a gun as one possible example. The answers do not say exactly what each person did. Survey questions Q63 and Q136
The outcome is the respondent’s own assessment. It does not tell us whether they actually called police, whether an offense had already occurred under applicable law, whether anyone was injured, or what would have happened without the gun. A respondent could believe an incident was over even if an investigator would classify it differently.
What happened in reported public locations?
Of the 2,627 reports, 2,284 took place inside the respondent’s own home or outside on their own property. 191 respondents selected a public location and 37 selected work. The remaining 115 selected someone else’s home or property or another location. The distinction matters when a statistic about all defensive use is brought into a debate about public carry. Public survey data, question Q64
Among the 191 public-location reports, 171 said the defender did not fire. Of those, 147 also said they believed no crime remained to report. These are public-location reports in a particular national sample. The survey gives no independently verified license status, state-law compliance, or public-carry exposure for these events. Nor does it tell us whether a rule in one location would have affected any of those incidents. Public survey data, questions Q63, Q64 and Q136
The public subgroup matters because defensive situations can arise away from home. Bruen protects public handgun carry for self-defense and invalidated New York’s special-need requirement. The survey describes experiences people reported. The legal scope of the right comes from the Court’s opinion. Supreme Court opinion
Why do news accounts show so many shots fired?
In a peer-reviewed study, David Hemenway, Chloe Shawah and Elizabeth Lites examined 418 defensive-use incidents from news accounts collected by the Gun Violence Archive. The incidents were from four months of 2019. The researchers found that the defender fired in about 90% of the cases. Their study describes the events that reached their news-based sample. In the NY Safe survey subset, 423 of 2,627 respondents, or 16.1%, said the defender fired. The two groups differ in year, selection, and incident definitions. The contrast illustrates how different sources can put very different encounters in front of the reader. Hemenway, Shawah and Lites, methods and results · Survey source file
Which records answer which question?
- Direct survey · English / NY Safe
- What people describe as defense, including uses without a shot or display. Recall and classification depend on the respondent.
- Victimization survey · Bureau of Justice Statistics
- How victims describe covered crimes and their responses, including crimes that never reached police. The questionnaire begins with victimization.
- News-based cases · Hemenway and colleagues
- Details of incidents that made the news. About 90% of the 418 analyzed cases involved a shot by the defender.
- Police records · NYPD and other agencies
- Incidents police learned about and recorded under their definitions. Shooting counts track recorded shootings.
A shooting count can describe shootings. It cannot tell us what happened in an encounter where the defender never fired.
The Centers for Disease Control and Prevention summarizes older defensive-use research on its public firearm injury page. The Bureau of Justice Statistics describes the federal victimization survey, which asks about covered victimizations whether or not they reached police. That makes the NCVS broader than police records. The crime screen still differs from the direct defensive-use questions in English’s survey. NY Safe’s existing CDC explainer covers the national frequency debate. Here the question is what people report happened during a defensive encounter. CDC firearm injury page · BJS survey methodology · NY Safe’s CDC estimates explainer
What police reports, prosecution decisions and bail data leave out
Imagine someone facing a violent attack. The attacker sees the defender’s gun, backs away and leaves. If nobody tells police and police never learn about it, the encounter is absent from police data. If nobody fired, it would also be absent from a shooting count. That is a real measurement problem when lawmakers use shooting and complaint totals to assess both the harms and the protective uses of firearms.
Gary Kleck and Marc Gertz asked whether police were told about or otherwise learned of defensive gun uses. In their national study published in 1995, the answer was yes for 64.2% of the incidents against people in their analysis. That leaves about 36% that respondents said never became known to police. It is older, self-reported national evidence, and it leaves today’s New York rate open. In our 2026 survey subset, 1,728 respondents reported no shot by the defender and believed there was no crime left to report. Their answer gives us a reason to ask about reporting. The survey item does not record whether police actually found out. Kleck and Gertz, Table 3, panel L
The broader reporting gap is also large. The Bureau of Justice Statistics found that police learned about approximately 48% of nonfatal violent victimizations in its 2024 national survey. In a separate NYC analysis of federal victimization data, the share of violent-crime victims who said they stayed silent because they thought police would not help rose from 7% in 1997 to 16% in 2021, using three-year moving averages. Those figures give us evidence that confidence can affect which crimes reach police. They describe broader victimization, so they cannot supply a correction factor for defensive gun use. BJS, Criminal Victimization, 2024, Table 4 · NYC victimization study, Figure 3a
- Police complaint. If nobody reports an incident and police never learn about it, it never enters the police record. Even a known incident may be classified differently from how a victim understood it.
- Prosecution. After an arrest, the district attorney can file charges, change them or decline to prosecute. A declination can keep an arrest out of court. The existing police complaint belongs to a separate count.
- Bail and pretrial release. In a filed case, the court decides what release conditions the law permits or requires. Charge and statutory circumstances matter. The original complaint and the eventual case outcome are separate from that release decision. New York Criminal Procedure Law § 510.10
What prosecutors decline to file
These differences show up in New York’s own data. In 2024, the Bronx DA declined prosecution after 25% of arrests classified as violent felonies, compared with 3% to 4% in each other borough. Case mix, available evidence and screening practices can all matter. If residents expect that even a serious complaint will go nowhere, some may decide it is pointless to report. That is a plausible concern worth testing directly. The victim survey asked about expected police help, and the arrest study measured what prosecutors actually did. Neither measured how many victims stayed silent because they expected a DA to decline their case. Data Collaborative for Justice, Exhibit 3.1c · NYC victimization study
What bail studies track after release
Bail reform also warrants scrutiny on its own terms. One study found a small increase in any pretrial re-arrest within six months, with larger increases among people the researchers classified as higher risk. A separate NYC systemwide study found no significant overall effect on re-arrest, though it found an increase in pretrial violent-felony re-arrest among defendants who already had a separate pending case. These studies measure re-arrest. They leave open whether a victim’s expectations about a release decision influence reporting. A serious safety review should track what happens after release and ask victims why they decided whether to come forward. Difference-in-differences bail study · NYC systemwide bail study
Can defensive gun use save lives?
Yes. Lawful defensive gun use can save lives. In July 2022, Elisjsha Dicken stopped an active shooter at Indiana’s Greenwood Park Mall. Greenwood police documented that the armed civilian stopped the shooter about 15 seconds after the gunman came out of the restroom. A congressional account says Dicken was carrying lawfully under Indiana law. Dicken ended an attack that was already killing people. That is what saving lives can look like. The exact number of people who would otherwise have been killed or injured is unknowable. Greenwood Police Department after-action report · U.S. Senate resolution announcement · Congressional account of lawful carry
A no-shot report tells us how the defender says the gun was used. It leaves the effect on injury open. A threat could have ended for several reasons, and a gun may be used in situations that were more dangerous to begin with. Researchers studying victim injury face a timing problem as well. An injury that occurred before the victim acted cannot be attributed to that action.
Jongyeon Tark and Gary Kleck studied 27,595 personal-contact incidents recorded by the federal National Crime Victimization Survey in 1992–2001. Their analysis found that several forms of resistance, including armed resistance, appeared to reduce injury risk compared with nonresistance. The authors cautioned that some estimates were unstable because the relevant cases were few. Tark and Kleck, Criminology
David Hemenway and Sara Solnick used a later NCVS window, 2007–2011. Among more than 14,000 victim-present incidents, 127 involved defensive gun use. The observed percentage injured after defensive gun use was 4.1%, versus 4.2% after any protective action. The “any action” group includes firearm defense. These are observed shares from that sample, and the authors found little evidence of a unique injury or property-loss advantage compared with other protective actions. A study comparing several actions with no resistance is asking a different question from a study testing whether the gun adds an advantage over other actions. Hemenway and Solnick, Preventive Medicine
How NY Safe counted the reports
We used the public Harvard Dataverse file 2026 National Firearms Survey (Wave 2), version 1.0, file 14168601. We selected finished responses that identified firearm ownership on Q2, gave a positive whole-number count in at least one of Q88_1, Q88_4 or Q88_5, answered yes to the lifetime-use question Q13, and selected exactly one encounter on Q15. That left 2,802 respondents. We then excluded 175 who selected the animal-threat outcome on Q136. All 2,627 remaining records had one of the three recognized Q63 action categories and one of the two non-animal outcomes on Q136. Q64 supplies the reported location. Dataset, version and raw file · Download the public survey file
The counts are unweighted. The denominator is the 2,627 included reports, not all adults, all firearm owners or all defensive uses. The no-shot share adds the “showed without firing” and “neither” rows, then divides 2,204 by 2,627. The combined outcome uses the 1,165 and 563 cells for those two rows, then divides 1,728 by 2,627. The 147 public-location reports use the same rule within the 191 records that selected public location. Responses describe past experiences at the time of a 2026 survey. They cannot establish a count for calendar year 2026.
The one-encounter rule gives each person one contribution. It leaves out respondents who reported repeat experiences. A wider count brings those people in, but it also gives more influence to respondents with more detailed entries. That is why we show the sensitivity results.
In their critique of English’s first wave, Deborah Azrael and colleagues argued that the survey design and analysis could not establish a reliable annual defensive-use count. English defended the sample and argued that weighting made little difference to his first-wave results. In our Wave 2 analysis, even a perfectly reproduced count cannot check whether someone recalled an encounter accurately or labeled their conduct correctly. Azrael and colleagues’ first-wave critique · English’s response
For a comparison of public warnings and later crime trends, see the October 2026 fact sheet from the Second Amendment Foundation, a gun-rights advocacy organization. Comparing dates can test a forecast of an immediate surge. Measuring a specific law’s effect takes a credible comparison with what would happen without it. SAF fact sheet
What changes when we count differently?
The share reporting no shot falls from 83.9% to 77.9% if we count only the one-encounter reports that describe displaying or firing the gun. Across all available non-animal entries, it is 78.3%. That broader count includes multiple entries from some people. These are descriptive checks on the public data, separate from English’s published 76.3%.
| Reports included | Count | No shot | No shot + perceived no crime |
|---|---|---|---|
| One encounter, non-animal | 2,627 | 83.9% | 65.8% |
| One encounter, include animal reports | 2,802 | 83.2% | 61.7% |
| One encounter, display or firing only | 1,916 | 77.9% | 60.8% |
| One listed encounter per respondent | 5,929 | 80.8% | 64.9% |
| All available non-animal entries | 10,520 | 78.3% | 61.3% |
| Repeat entries only | 7,893 | 76.5% | 59.9% |
For the one-per-person comparison, we used the first listed entry for repeat respondents and excluded it if it described an animal threat. The broader count uses up to five detailed entries per person as supplied by the survey. It leaves out lifetime encounters that the questionnaire did not detail. The 191 public-location reports are a small descriptive subgroup. We found no usable weight field in the deposited file. We report exact counts. A conventional margin of error would assume a sampling design this analysis cannot establish.
Dataverse variable metadata · Download the TSV and its questionnaire headers
What a safer New York should make possible
I believe New York is better off when more eligible, responsible adults have a meaningful opportunity to lawfully own and carry a firearm for self-defense. When a person lawfully protects themselves or someone else from violence, that protection counts as a public-safety benefit. The case for protecting that option is strongest when we take the risk of misuse just as seriously as the potential benefit. Clear and timely licensing, safe storage, sound judgment, and enforcement against actual violence all belong in that approach. New York’s licensing statute
The strongest objection deserves a direct answer. RAND’s research review finds supportive evidence that shall-issue carry laws can increase homicide and violent crime. A working paper by John Donohue and colleagues illustrates how the two effects might coexist. Its incident analysis suggested that defensive gun use could reduce injury or death for the victim in the cases studied. Its broader analysis associated higher levels of defensive use with more violence overall. That working paper uses selected incident records and a proxy for carry. Both the individual benefit and potential community costs belong in the evaluation of any specific New York proposal. RAND’s review of concealed-carry laws · Donohue and colleagues’ working paper
Where access can increase harm
Access also brings risks that our survey table cannot measure. A California cohort study found a higher firearm suicide rate among handgun owners. A separate cohort found higher homicide rates among adults living with handgun owners, including higher rates of fatal shootings by intimate partners. These studies followed different California populations and measured different outcomes from the defense survey. California firearm suicide cohort · California cohabitant homicide cohort
CDC documented fatal unintentional shootings involving children and firearms stored unlocked or loaded. ATF tracks firearms stolen from private owners. A gun can also escalate a confrontation if the threat is misread. Responsible ownership includes secure storage, planning for unauthorized access and training that teaches when to disengage. CDC child firearm injury report · ATF firearm theft report · NY Safe’s judgment guide
What Albany should measure and publish
- Lawful access. Publish active licenses, new issues, renewals and wait times by county and license type. Keep personal records private.
- What happens to reported crime. Publish separately defined counts of complaints, arrests, DA declinations, pretrial release decisions and case outcomes.
- What never reaches police. Use a confidential victim survey to ask whether police learned of the encounter and why someone did or did not report it.
- What each restriction accomplishes. Show the harm the rule is intended to reduce, test whether it does so and account for any lawful defensive options it forecloses.
Frequently asked questions
Did lawful carry cause New York City shootings to fall?
City shootings fell as NYC approved thousands of carry applicants after Bruen. Those records establish that the two developments coexisted. Isolating the effect of legal carry would require information on actual carry and a credible comparison with what would have happened under different rules. NYC carry approval reporting · NYPD incident data
Can defensive gun use occur without firing?
Yes, under the definition used by this survey. It asked separately about firing, displaying, and neither. In NY Safe’s defined subset, 2,204 of 2,627 respondents reported that the defender did not fire. The survey label describes an answer to a research question.
What does “no crime remained to report” mean?
It means the respondent believed the firearm use had deterred the conduct to the point that there was no crime left to report. The question asks for their judgment. It does not record actual calls to police, a charging decision or a court’s ruling.
Does the public-location result apply to New York license holders?
The 191 reports were from a national survey subset. The public-location answer alone cannot identify New York incidents, carry-license status at the time, or whether a firearm was possessed lawfully in that setting.
Is showing a gun without firing legal in New York?
The circumstances matter. Penal Law § 120.14(1) covers intentionally placing or attempting to place someone in reasonable fear of injury or death by displaying a firearm or an apparent firearm. A claim of justification requires its own factual and legal analysis. If you have a question about a specific event, consult a qualified New York attorney. Menacing statute · New York Penal Law § 35.15
What judgment training should cover
A useful licensing class should help you explain your decision in plain English. What facts made the threat immediate? What safer option was available? What changed when the other person backed away? These questions belong alongside safe handling and live-fire qualification. The point is to recognize where an encounter can end safely before choices narrow.
You can see how NY Safe approaches that work in our concealed-carry judgment guide. If you are comparing training providers, look at what they actually teach about avoidance, de-escalation, New York’s use-of-force framework and firearms safety. Then ask how students get to practice and receive feedback on those decisions.
See what NY Safe’s 16+2 course covers
NY Safe’s New York licensing course pairs de-escalation and use-of-force instruction with controlled simulator scenarios, a written proficiency test and live-fire qualification. The course overview shows what you’ll study and how to find upcoming dates.
Next Available Classes
Upcoming New York 16+2 CCW Class Dates
Limited to 15 students per class. Seats fill quickly.
No responses yet