Defensive Gun Use Kleck CDC Study
courtesy and Getty
Previous Post
Next Post

By Robert B. Young, MD

In April, we wrote about the unpublished Centers for Disease Control survey from 1996-1998 that evidently confirmed the findings of Gary Kleck, PhD and Marc Gertz, PhD from 1992 that there were likely about 2.4 million civilian defensive gun uses (DGUs) each year in the United States. Dr. Kleck discovered that CDC data last fall and, extrapolating it to compare that data from 13 states to his own national survey, concluded that the results fell within about 2.5% of each other.

It may be worth reviewing our April piece to understand the background for Kleck’s re-analysis. Kleck and Gertz’ original survey done in 1995 is called the National Self-Defense Survey. The CDC’s work was contained in its annual Behavioral Risk Factor Surveillance System survey.

Robert VerBruggen of National Review reasonably criticized Kleck’s conclusion on the basis that the 13 states sampled over three years by the CDC in the BRFSS may not be directly comparable to the national sample in the NSDS (see the updated introduction to Reason’s April 20 article). Kleck withdrew his original paper in order to go over the data further.

He has now published his own update, which can be found here (requiring free registration to access the full paper). Reason has also published its discussion of Kleck’s “Second Look”.

Kleck revised his conclusion about the BRFSS that it substantiates in itself an average of 1.1 million DGUs annually during the three years it asked about them. It’s clear that with surveys from these few states (none sampled in all three years), there could be substantial variation in their ability to represent the nation as a whole. There could also be substantial differences should they be queried in a year particularly high or low in DGU occurrences.

There were also a significant number of respondents who either refused to answer or said they didn’t know whether they had a DGU. One suspects that many of those respondents really had, but did not want to go on record. Maybe a few were unsure: in 90% of DGUs, no shot even is fired–the assailant is scared off.

But all in all, each year of the BRFSS answers led Kleck to varying estimates of total DGUs nationally: 620,648 in 1996, 854,498 in 1997, and 1,940,455 in 1998. For what it’s worth, violent crime was on a steep decline during those three years—not that increasing civilian DGUs necessarily contributed.

The BRFSS survey questions were optional for each state, which would explain why so few were represented. If CDC had really wanted to check Kleck and Gertz’ 1995 conclusions, it would have asked them in a national survey. But as we and many others speculate, it is likely that CDC, at the time a hotbed of anti-gun sentiment, did this as a trial study to test whether their findings could refute the NSDS results. Within the inherent limits of the BRFSS, it is clear that it confirmed that there are far more DGUs each year than any gun control theoretician wants to believe.

Consider that in the 1992 NSDS 46% of those reporting DGUs believed someone “might have”, “probably would have”, or “almost certainly would have” been killed otherwise.  Even of 1.1 million DGUs, nearly half may have saved a life (and we ought to assume, conservatively, that at least the 16% “almost certainly” did).

There are roughly 100,000 people shot in the United States annually, and something over 30,000 die. If this 1/3 vs. 2/3 ratio of deaths to injuries in actual shootings pertains in these DGUs, that makes for at least 176,000 lives saved—less some attackers who lost their lives to defenders. This enormous benefit dwarfs — both in human and economic terms — the losses trumpeted by hoplophobes who only choose to see the risk side of the equation.

Kleck is still correct, whether precisely or not. Remember that the NSDS was more comprehensive and sampled the national population for one year compared to the BRFSS sampling some states inconsistently. That makes Kleck and Gertz’ work the more valid of the two in any case.

Late 1990’s CDC leadership must have been scared off by their own numbers. However you slice it, there are tremendous numbers of DGUs done by Americans who take self-defense seriously and effectively. Thank goodness and the Second Amendment!

Reality bites those who deny it. Facing it squarely head on beats sticking one’s head in the sand and mooning it.

Kleck and Gertz were right, now more clearly than ever. Their critics never were.

This case is still closed.


DRGO Editor Robert B. Young, MD is a psychiatrist practicing in Pittsford, NY, an associate clinical professor at the University of Rochester School of Medicine, and a Distinguished Life Fellow of the American Psychiatric Association.

This article originally appeared at and is reprinted here with permission. 

Previous Post
Next Post


  1. Guns still kill fewer people than doctors do, and they can be used for self defense…🤔🤔🤔🤔🤔

  2. Regarding: “There are roughly 100,000 people shot in the United States annually, and something over 30,000 die.”; it would make the argument more solid if the suicide numbers were removed, if the data allow. The relationship between DGUs and suicide attempts is tenuous, at best and the 1/3-2/3 ratio may not apply.

    I can see why the CDC abandoned a wider survey.

    • I know!! I saw that about combining suicide data in with DGUs. Don’t believe anything from both the pro and anti gun sides, look for primary data sources.

      • Suicides definitely shouldn’t be included. It’s going to happen whether there’s a gun, knife, rope or pills.

        My brother had a gun in his apartment he was renting and chose to hang himself. This was just a few weeks after Robin Williams. I don’t know if the choice to go that way was coincidental or not.

    • From this we can conclude that 70,000 people are shot with 9mm, and the remaining 30,000 shot with .45.

    • while I don’t doubt the veracity of this data…it has yet to happen to anyone I know who carries…thankfully….

  3. More importantly, the National Institute of Justice and the CDC funded the gun control advocate, David Hemenway, to do his own DGU survey in the wake of Kleck’s work. They did this knowing that Hemenway would give them the results they wanted. That, to me, is the bombshell finding. I can’t wait until it makes the cover the New York Times and Congress holds hearings over this.

    Oh wait, I must have dozed off and been dreaming.

    • Hemenway is a full on nut. He 100% refused to control for criminal violence. In fact his taxpayer funded “studies”allegedly showing owning a gun increased risk to household members are definitive junk science. Over 90% of murder is criminal on criminal with illegally possessed guns.

      If you don’t have a meth lab, are not dealing drugs, a pimp, a robber, a gang member etc, ie in the 6% of the population who is 90% of murder victims, but are int he 94% of the population that is not a serious criminal, then having a gun makes you and household members safer than not having one.

      Hemengways methodology would be like claiming that simply living in Oregon “causes” you to have lung cancer and moving to Utah stops you from getting it,. Any epidemiologist would tell you it is specious not considering the smoking rate in Oregon bing almost twice as high as Utah. Smoking is the overwhelming driving risk with lung cancer, just as criminal activity is the overwhelming driving risk with gunshot victimization.

      The vast majority of Oregon residents who are not smokers have no elevated risk of lung cancer, and the vast majority of gun owners who are not criminals don’t have more dangerous households due to gun ownership — but safer ones.

  4. This is common knowledge. Everyone with half a brain knows guns save more lives annually by 10x than are killed by them. Especially considering 75% of the gun deaths yearly in this country are suicides. And if someone is bound and determined to kill themselves they damned sure don’t need a gun to do it. I think the actual legitimate murders committed yearly in this country is closer to 1200 to 1500. Because I don’t consider the 7000 gang related deaths legitimate because they’re basically animals living in a law of the fittest society. Easily other forms of murder outnumber gun deaths, fists, knifes hammers etc. The way the CDC skews their numbers is outlandish, they also lump self defense and killings by cops while using a gun as illegitimate gun deaths to pad their numbers to please the liberals.

Comments are closed.