The debate over how often people use guns to protect themselves in the U.S. continues to smolder.
Gary Kleck, a professor of criminology at Florida State University who is a controversial figure for gun control advocates, recently published a paper examining three surveys in the 1990s conducted by the Centers for Disease Control and Prevention (CDC), entitled, “What Do CDC Surveys Say About the Frequency Of Defensive Gun Uses?”
The paper looks at a large portion of the literature on defensive gun use(DGU) in the US. It focuses on data from three annual surveys in 1996, 1997 and 1998, produced by the CDC’s Behavioral Risk Factor Surveillance System (BRFSS), which by Kleck’s calculations, show an average of 1,138,534 instances of DGUs per year.
This number is far higher than the number commonly cited from the National Crime Victimization Survey (NCVS), which reports annual estimates of roughly 60,000 instances of DGU.
Kleck writes, “CDC’s survey data confirm previous high estimates of DGU prevalence, disconfirm estimates derived from the National Crime Victimization Survey, and indicate that defensive uses of guns by crime victims are far more common than offensive uses by criminals.”
The 1996 survey asked people in six states (AK, KY, LA, MD, NH, WV), the 1997 survey asked people in seven states (CO, HA, MS, NH, NJ, ND, OH), and the 1998 survey asked people in four states (LA, MT, NJ, PA).
There is a high degree of variance among the nearly two-dozen studies compared in the paper. For example, another study shown in Kleck’s paper by the CDC in 1994 found an implied annual estimate of three million DGUs, half a million more than Kleck himself and a colleague, Marc Gertz, found in a survey study they did in 1993.
However, many of the studies, while having vastly different results, had different recall periods. For example, a recent 2017 Pew survey found there were 2.6 million instances of DGU, but there was no limit on the recall period. This number would invariably drop considerably had the recall period been relegated to instances of DGU within the past year.
Kleck writes that the surveys by the BRFSS are of high quality, and ask more people (3,197-4,500 adults) than any other surveys except the above-mentioned one conducted by Kleck and Gertz in 1993. It should be mentioned that while Kleck does not list the NCVS in his direct comparisons, it is by far the most comprehensive survey, conducted by the Bureau of Justice Statistics, asking roughly 90,000 people in 1994.
The question posed to respondents in the BRFSS surveys was: “During the last 12 months, have you confronted another person with a firearm, even if you did not fire it, to protect yourself, your property, or someone else?”
Kleck notes that the question was well-worded and gave respondents little room for error or misinterpretation. It included the use of any type of firearm, asked about a specific recall period, and asked respondents to report the incident even if they didn’t fire the gun.
Protection is one of the main reasons people cite for owning a gun, so the substantial amount of DGUs found by the BRFSS according Kleck’s calculations can serve to support that rationale.
False positives and false negatives, in which respondents lie or unintentionally give the wrong answer, can be a significant issue for DGU studies. Because of the fact that some portion of defensive gun use is likely illegal, for example, using it away from the home without a permit, respondents may not wish to divulge that information. That is just one way in which respondents may not give accurate answers to a very difficult question.
Kleck cites this as one of the reasons the NCVS may be underrepresenting the amount of DGUs each year in the US. This is because they are asked about the incident after being asked where the incident occurred.
David Hemenway, Director of the Harvard Injury Control Research Center, has been a vocal opponent of Kleck’s and vice-versa since the 1990s. While Kleck states that there is more likelihood people will give false negatives than false positives, Hemenway says that false positives are a reasonable possibility from people who are fearful of gun control measures.
Hemenway wrote, “A few might actually deliberately lie on a telephone survey to help boost the numbers for the sake of their political beliefs concerning the dangers of gun control,” in a 1997 paper called, “Survey Research and Self-Defense Gun Use: An Explanation of Extreme Overestimates.
Hemenway and others have written that surveys can suffer from the problem of “telescoping,” in which respondents give answers that are true but happened outside the timeframe of the question. Hemenway writes that the NCVS compensates for this with follow up interviews every six months, over a three year period. “Analysis of the NCVS unbounded (or first-time) panel in comparison with the second indicates a substantial amount of telescoping of criminal victimization.”
Phillip Cook, a criminologist at the Sanford School of Public Policy at Duke University, has also written critically of Kleck’s work in a 2015 paper called, “Elusive Facts About Gun Violence: Where Good Surveys Go Bad.” Cook states that Kleck’s findings of 2.5 million DGUs per year is more than double the total number of robberies and assaults committed with a gun, “which in turn is far more than the number of gun crimes known to the police.”
A National Research Council report also finds that Kleck’s estimates appear exaggerated and says that it is almost certain that “some of what respondents designate as their own self-defense would be construed as aggression by others.”
The point is often made that official estimates from law enforcement agencies would provide more accurate data as to how many DGUs a year there actually are. Kleck argues, however, that “there is currently no feasible way to measure the prevalence of DGU other than surveys. “
He continues: “Certainly police data cannot provide adequate estimates given the unwillingness of most crime victims to even report their victimizations to the police (U.S. Bureau of Justice Statistics, 1999), never mind the controversial fact that they had threatened or attacked another person with a firearm.”
Cook also responds to a Kleck and Gertz finding of 200,000 people actually shooting their assailants per year, which he notes is again double the number of people killed or treated in emergency room departments.
Cook cites a series of jail surveys found that a high percentage of inmates said they had indeed been shot, “had the scars to prove it, (and) more than 90 percent of those who had been shot reported that they had indeed been treated in a hospital.”
Cook also mentions a study by three researchers at the Harvard School of Public Health, including Hemenway, which found that survey respondents who were asked about both DGUs and victimization by guns, reported being victimized substantially more than DGUs.
A look at the literature underlines the fact that this is a controversial area of gun research. Studies like the one done by Kleck and Gertz clearly lie on the high end of the spectrum, while the NCVS and Hemenway studies lie on the low end of that spectrum. It would seem to be a fair assumption that the actual number of DGUs annually are somewhere in between the extremes.
Nevertheless, the debate highlights the fact that DGU is an extremely difficult statistic to measure, and that there is a serious need for more research on gun violence in general in the US so that more of a consensus can be reached.
Adding to the debate about his assertions, Kleck claimed that the CDC never actually reported the data, although it is available for anyone who knows where to look.
The CDC did not immediately respond to a request for comment.
Dane Stallone is a TCR News Intern. He welcomes reader’s comments.