University Study Confirms Private Firearms

Stop Crime 2.5 Million Times Each Year

By J. Neil Schulman

Gary Kleck, Ph.D. is a professor in the School of Criminology and Criminal Justice at Florida State University in Tallahassee and author of “Point Blank: Guns and Violence in America” (Aldine de Gruyter, 1991), a book widely cited in the national gun-control debate. In an exclusive interview, Dr. Kleck revealed some preliminary results of the National Self- Defense Survey which he and his colleague Dr. Marc Gertz conducted in Spring, 1993. Though he stresses that the results of the survey are preliminary and subject to future revision, Kleck is satisfied that the survey’s results confirm his analysis of previous surveys which show that American civilians commonly use their privately-owned firearms to defend themselves against criminal attacks, and that such defensive uses significantly outnumber the criminal uses of firearms in America. The new survey, conducted by random telephone sampling of 4,978 households in all the states except Alaska and Hawaii, yield results indicating that American civilians use their firearms as often as 2.5 million times every year defending against a confrontation with a criminal, and that handguns alone account for up to 1.9 million defenses per year. Previous surveys, in Kleck’s analysis, had underrepresented the extent of private firearms defenses because the questions asked failed to account for the possibility that a particular respondent might have had to use his or her firearm more than once.

Dr. Kleck will first present his survey results at an upcoming meeting of the American Society of Criminology, but he agreed to discuss his preliminary analysis, even though it is uncustomary to do so in advance of complete peer review, because of the great extent which his earlier work is being quoted in public debates on firearms public policy.
The interview was conducted September 14-17, 1993 by J. Neil Schulman, a novelist, screenwriter, and journalist who has written extensively on firearms public policy for several years.

Readers may be interested to know that Kleck is a member of the ACLU, Amnesty International USA, and Common Cause, among other politically liberal organizations. He is also a lifelong registered Democrat. He is not and has never been a member of or contributor to the NRA, Handgun Control Inc., or any other advocacy group on either side of the gun-control issue, nor has he received funding for research from any such organization.


SCHULMAN: Dr. Kleck, can you tell me generally what was discovered in your recent survey that wasn’t previously known?

KLECK: Well, the survey mostly generated results pretty consistent with those of a dozen previous surveys which generally indicates that defensive use of guns is pretty common and probably more common than criminal uses of guns. This survey went beyond previous ones in that it provided detail about how often people who had used a gun had done so. We asked people was the gun used defensively in the past five years and if so how many times did that happen and we asked details about what exactly happened. We nailed down that each use being reported was a bona fide defensive use against a human being in connection with a crime where there was an actual confrontation between victim and offender. Previous surveys were a little hazy on the details of exactly what was being reported as a defensive gun use. It wasn’t, for example, clear that the respondents weren’t reporting investigating a suspicious noise in their back yard with a gun where there was, in fact, nobody there. Our results ended up indicating, depending on which figures you prefer to use, anywhere from 800,000 on up to 2.4, 2.5 million defensive uses of guns against human beings — not against animals — by civilians each year.

SCHULMAN: Okay. Let’s see if we can pin down some of these figures. I understand you asked questions having to do with just the previous one year. Is that correct?

KLECK: That’s correct. We asked both for recollections about the preceding five years and for just what happened in the previous one year, the idea being that people would be able to remember more completely what had happened just in the past year.

SCHULMAN: And your figures reflect this?

KLECK: Yes. The estimates are considerably higher if they’re based on people’s presumably more-complete recollection of just what happened in the previous year.

SCHULMAN: Okay. So you’ve given us the definition of what a “defense” is. It has to be an actual confrontation against a human being attempting a crime? Is that correct?

KLECK: Correct.

SCHULMAN: And it excludes all police, security guards, and military personnel?

KLECK: That’s correct.

SCHULMAN: Okay. Let’s ask the “one year” question since you say that’s based on better recollections. In the last year how many people who responded to the questionnaire said that they had used a firearm to defend themselves against an actual confrontation from a human being attempting a crime?

KLECK: Well, as a percentage it’s 1.33 percent of the respondents. When you extrapolate that to the general population, it works out to be 2.4 million defensive uses of guns of some kind — not just handguns but any kind of a gun — within that previous year, which would have been roughly from Spring of 1992 through Spring of 1993.

SCHULMAN: And if you focus solely on handguns?

KLECK: It’s about 1.9 million, based on personal, individual recollections.

SCHULMAN: And what percentage of the respondents is that? Just handguns?

KLECK: That would be 1.03 percent.

SCHULMAN: How many respondents did you have total?

KLECK: We had a total of 4,978 completed interviews, that is, where we had a response on the key question of whether or not there had been a defensive gun use.

SCHULMAN: So roughly 50 people out of 5000 responded that in the last year they had had to use their firearms in an actual confrontation against a human being attempting a crime?

KLECK: Handguns, yes.

SCHULMAN: Had used a handgun. And slightly more than that had used any gun.

KLECK: Right.

SCHULMAN: So that would be maybe 55, 56 people?

KLECK: Something like that, yeah.

SCHULMAN: Okay. I can just hear critics saying that 50 or 55 people responding that they used their gun and you’re projecting it out to figures of around 2 million, 2-1/2 million gun defenses. Why is that statistically valid?

KLECK: Well, that’s one reason why we also had a five-year recollection period. We get a much larger raw number of people saying, “Yes, I had a defensive use.” It doesn’t work out to be as many per year because people are presumably not remembering as completely, but the raw numbers of people who remember some kind of defensive use over the previous five years, that worked out to be on the order of 200 sample cases. So it’s really a small raw number only if you limit your attention to those who are reporting an incident just in the previous year. Statistically, it’s strictly the raw numbers that are relevant to the issue.

SCHULMAN: So if between 1 percent to 1-1/3 percent of your respondents are saying that they defended themselves with a gun, how does this compare, for example, to the number of people who would respond that they had suffered from a crime during that period?

KLECK: I really couldn’t say. We didn’t ask that and I don’t think there are really any comparable figures. You could look at the National Crime Surveys for relatively recent years and I guess you could take the share of the population that had been the victims of some kind of violent crime because most of these apparently are responses to violent crimes. Ummm, let’s see. The latest year for which I have any data, 1991, would be about 9 percent of the population had suffered a personal crime — that’s a crime with personal contact. And so, to say that 1 percent of the population had defended themselves with a handgun is obviously still well within what you would expect based on the share of the population that had suffered a personal crime of some kind. Plus a number of these defensive uses were against burglars, which isn’t considered a personal crime according to the National Crime Survey. But you can add in maybe another 5 percent who’d been a victim of a household burglary.

SCHULMAN: Let’s break down some of these gun defenses if we can. How many are against armed robbers? How many are against burglars? How many are against people committing a rape or an assault?

KLECK: About 8 percent of the defensive uses involved a sexual crime such as an attempted sexual assault. About 29 percent involved some sort of assault other than sexual assault. Thirty-three percent involved a burglary or some other theft at home. Twenty-two percent involved robbery. Sixteen percent involved trespassing. Note that some incidents could involve more than one crime.

SCHULMAN: Do you have a breakdown of how many occurred on somebody’s property and how many occurred, let’s say, off somebody’s property where somebody would have had to have been carrying a gun with them on their person or in their car?

KLECK: Yes. We asked where the incident took place. Seventy-two percent took place in or near the home, where the gun wouldn’t have to be “carried” in a legal sense. And then some of the remainder, maybe another 4 percent, occurred in a friend’s home where that might not necessarily involve carrying. Also, some of these incidents may have occurred in a vehicle in a parking lot and that’s another 4 percent or so. So some of those incidents may have involved a less-regulated kind of carrying. In many states, for example, it doesn’t require a license to carry a gun in your vehicle so I’d say that the share that involved carrying in a legal sense is probably less than a quarter of the incidents. I won’t commit myself to anything more than that because we don’t have the specifics of whether or not some of these away-from-home incidents occurred while a person was in a car.

SCHULMAN: All right. Well, does that mean that approximately a half million times a year somebody carrying a gun away from home uses it to defend himself or herself?

KLECK: That’s what it would imply, yes.

SCHULMAN: All right. As many as one-half million times every year somebody carrying a gun away from home defends himself or herself.

KLECK: Yes, about that. It could be as high as that. I have many different estimates and some of the estimates are deliberately more conservative in that they exclude from our sample any cases where it was not absolutely clear that there was a genuine defensive gun use being reported.

SCHULMAN: Were any of these gun uses done by anyone under the age of 21 or under the age of 18?

KLECK: Well we don’t have any coverage of persons under the age of 18. Like most national surveys we cover only adults age 18 and up.

SCHULMAN: Did you have any between the ages of 18 and 21?

KLECK: I haven’t analyzed the cross tabulation of age with defensive gun use so I couldn’t say at this point.

SCHULMAN: Okay. Was this survey representative just of Florida or is it representative of the entire United States?

KLECK: It’s representative of the lower 48 states.

SCHULMAN: And that means that there was calling throughout all the different states?

KLECK: Yes, except Alaska and Hawaii, and that’s also standard practice for national surveys; because of the expense they usually aren’t contacted.

SCHULMAN: How do these surveys make their choices, for example, between high-crime urban areas and less-crime rural areas?

KLECK: Well, there isn’t a choice made in that sense. It’s a telephone survey and the telephone numbers are randomly chosen by computer so that it works out that every residential telephone number in the lower 48 states had an equal chance of being picked, except that we deliberately oversampled from the South and the West and then adjusted after the fact for that overrepresentation. It results in no biasing. The results are representative of the entire United States, but it yields a larger number of sample cases of defensive gun uses. They are, however, weighted back down so that they properly represent the correct percent of the population that’s had a defensive gun use.

SCHULMAN: Why is it that the results of your survey are so counter-intuitive compared to police experience?

KLECK: For starters, there are substantial reasons for people not to report defensive gun uses to the police or, for that matter, even to interviewers working for researchers like me — the reason simply being that a lot of the times people either don’t know whether their defensive act was legal or even if they think that was legal, they’re not sure that possessing a gun at that particular place and time was legal. They may have a gun that’s supposed to be registered and it’s not or maybe it’s totally legally owned but they’re not supposed to be walking around on the streets with it.

SCHULMAN: Did your survey ask the question of whether people carrying guns had licenses to do so?

KLECK: No, we did not. We thought that would be way too sensitive a question to ask people.

SCHULMAN: Okay. Let’s talk about how the guns were actually used in order to accomplish the defense. How many people, for example, had to merely show the gun, as opposed to how many had to fire a warning shot, as to how many actually had to attempt to shoot or shoot their attacker?

KLECK: We got all of the details about everything that people could have done with a gun from as mild an action as merely verbally referring to the gun on up to actually shooting somebody.

SCHULMAN: Could you give me the percentages?

KLECK: Yes. You have to keep in mind that it’s quite possible for people to have done more than one of these things since they could obviously both verbally refer to the gun and point it at somebody or even shoot it.


KLECK: Fifty-four percent of the defensive gun uses involved somebody verbally referring to the gun. Forty-seven percent involved the gun being pointed at the criminal. Twenty-two percent involved the gun being fired. Fourteen percent involved the gun being fired at somebody, meaning it wasn’t just a warning shot; the defender was trying to shoot the criminal. Whether they succeeded or not is another matter but they were trying to shoot a criminal. And then in 8 percent they actually did wound or kill the offender.

SCHULMAN: In 8 percent, wounded or killed. You don’t have it broken down beyond that?

KLECK: Wound versus kill? No. Again that was thought to be too sensitive a question. Although we did have, I think, two people who freely offered the information that they had, indeed, killed someone. Keep in mind that the 8 percent figure is based on so few cases that you have to interpret it with great caution.

SCHULMAN: Did anybody respond to a question asking whether they had used the gun and it was found afterward to be unjustified?

KLECK: We did not ask them that question although we did ask them what crime they thought was being committed. So in each case the only incidents we were accepting as bona fide defensive gun uses were ones where the defender believed that, indeed, a crime had been committed against them.

SCHULMAN: Did you ask any follow-up questions about how many people had been arrested or captured as a result of their actions?


SCHULMAN: Did you ask any questions about aid in law enforcement, such as somebody helps a police officer who’s not themselves an officer?

KLECK: No. I imagine that would be far too rare an incident to get any meaningful information out of it. Highly unlikely that any significant share of these involved assisting law enforcement.

SCHULMAN: The question which this all comes down to is that we already have some idea, for example from surveys on CCW license holders, how rare it is for a CCW holder to misuse their gun in a way to injure somebody improperly. But does this give us any idea of what the percentages are of people who carry a gun having to use it in order to defend himself or herself? In other words, comparing the percentage of defending yourself to the percentage of being attacked, does this tell us anything?

KLECK: We asked them whether they carried guns at any time but we didn’t directly ask them if they were carrying guns, in the legal sense, at the time they had used their gun defensively. So we can probably say what fraction of gun carriers in our sample had used a gun defensively but we can’t say whether they did it while carrying. They may, for example, have been people who at least occasionally carried a gun for protection but they used a gun defensively in their own home.

SCHULMAN: So what percentage of gun carriers used it defensively?

KLECK: I haven’t calculated it yet so I couldn’t say.

SCHULMAN: So if we assume, let’s say, that every year approximately 9 percent of people are going to be attacked, and approximately every year that 1 percent of respondents used their guns to defend against an attack, is it fair to say that around one out of nine people attacked used their guns to defend themselves?

KLECK: That “risk of being attacked” shouldn’t be phrased that way. It’s the risk of being the victim of a personal crime. In other words, it involved interpersonal contact. That could be something like a nonviolent crime like purse snatching or pickpocketing as well. The fact that personal contact is involved means there’s an opportunity to defend against it using a gun; it doesn’t necessarily mean there was an attack on the victim.

SCHULMAN: Did you get any data on how the attackers were armed during these incidents?

KLECK: Yes. We also asked whether the offender was armed. The offender was armed in 47.2 percent of the cases and they had a handgun in about 13.6 percent of all the cases and some other kind of gun in 4.5 percent of all the cases.

SCHULMAN: So in other words, in about a sixth of the cases, the person attacking was armed with a firearm.

KLECK: That’s correct.

SCHULMAN: Okay. And the remainder?

KLECK: Armed with a knife: 18.1 percent, 2 percent with some other sharp object, 10.1 percent with a blunt object, and 6 percent with some other weapon. Keep in mind when adding this up that offenders could have had more than one weapon.

SCHULMAN: So in approximately five sixths of the cases somebody carrying a gun for defensive reasons would find themselves defending themselves either against an unarmed attacker or an attacker with a lesser weapon?

KLECK: Right. About five-sixths of the time.

SCHULMAN: And about one-sixth of the time they would find themselves up against somebody who’s armed with a firearm.

KLECK: Well, certainly in this sample of incidents that was the case.

SCHULMAN: Which you believe is representative.

KLECK: It’s representative of what’s happened in the last five years. Whether or not it would be true in the future we couldn’t say for sure.

SCHULMAN: Are there any other results coming out of this which are surprising to you?

KLECK: About the only thing which was surprising is how often people had actually wounded someone in the incident. Previous surveys didn’t have very many sample cases so you couldn’t get into the details much but some evidence had suggested that a relatively small share of incidents involved the gun inflicting wounds so it was surprising to me that quite so many defenders had used a gun that way.

SCHULMAN: Dr. Kleck, is there anything else you’d like to say at this time about the results of your survey and your continuing analysis of them?

KLECK: Nope.

SCHULMAN: Then thank you very much.

KLECK: You’re welcome.

Reply to –
J. Neil Schulman
Mail: P.O. Box 94, Long Beach, CA
90801-0094 JNS
BBS: 1-310-839-7653,,,,25