Public Health and Gun Control: A Review
Part I: The Benefits of Firearms
I have related previously (Medical Sentinel, Spring and Summer 1997) how the 1991 American Medical Associationıs (AMA) campaign against domestic violence launched for public relation consumption went hand in hand with the public health establishmentıs 1979 stated objective of eradication of handguns in America, beginning with a 25 percent reduction by the year 2000.1-2 Towards that objective, in the 1980s, hundreds of articles describing politicized, biased, result-oriented research funded at taxpayers expense were published in the medical journals.3 One of the principle investigators was Dr. Arthur Kellermann, who now heads the Emory University School of Public Health.
A significant portion of the gun control agenda, not only of the public health but the entire health advocacy establishment, in fact, comes from Dr. Kellermann's landmark articles, particularly "Gun Ownership As a Risk Factor for Homicide in the Home," published in The New England Journal of Medicine (NEJM) in 1993.4 And yet, much of the methodology, not to mention conclusions in the article have been questioned by numerous investigators.5-7
Since at least the mid-1980s, Dr. Kellermann (and associates), whose work had been heavily-funded by the CDC, published a series of studies purporting to show that persons who keep guns in the home are more likely to be victims of homicide than those who donıt. In a 1986 NEJM paper, Dr. Kellermann and associates, for example, claimed their "scientific research" proved that defending oneself or oneıs family with a firearm in the home is dangerous and counter productive, claiming "a gun owner is 43 times more likely to kill a family member than an intruder."8
In a critical review and now classic article published in the March 1994 issue of the Journal of the Medical Association of Georgia (JMAG), Dr. Edgar Suter, Chairman of Doctors for Integrity in Policy Research (DIPR), found evidence of "methodologic and conceptual errors," such as prejudicially truncated data and the listing of "the correct methodology which was described but never used by the authors."5 Moreover, the gun control researchers failed to consider and underestimated the protective benefits of guns. Dr. Suter writes: "The true measure of the protective benefits of guns are the lives and medical costs saved, the injuries prevented, and the property protected not the burglar or rapist body count. Since only 0.1 - 0.2 percent of defensive uses of guns involve the death of the criminal, any study, such as this, that counts criminal deaths as the only measure of the protective benefits of guns will expectedly underestimate the benefits of firearms by a factor of 500 to 1,000."5
In 1993, in his landmark and much cited NEJM article (and the research, again, heavily funded by the CDC), Dr. Kellermann attempted to show again that guns in the home are a greater risk to the victims than to the assailants.4 Despite valid criticisms by reputable scholars of his previous works (including the 1986 study), Dr. Kellermann ignored the criticisms and again used the same methodology. He also used study populations with disproportionately high rates of serious psychosocial dysfunction from three selected state counties, known to be unrepresentative of the general U.S. population. For example, 53 percent of the case subjects had a history of a household member being arrested, 31 percent had a household history of illicit drug use, 32 percent had a household member hit or hurt in a family fight, and 17 percent had a family member hurt so seriously in a domestic altercation that prompt medical attention was required. Moreover, both the case studies and control groups in this analysis had a very high incidence of financial instability. In fact, in this study, gun ownership, the supposedly high risk factor for homicide was not one of the most strongly associated factors for being murdered. Drinking, illicit drugs, living alone, history of family violence, living in a rented home were all greater individual risk factors for being murdered than a gun in the home. One must conclude there is no basis to apply the conclusions of this study to the general population.
All of these are factors that, as Dr. Suter pointed out, "would expectedly be associated with higher rates of violence and homicide."5 It goes without saying, the results of such a study on gun homicides, selecting this sort of unrepresentative population sample, nullify the authors' generalizations, and their preordained, conclusions can not be extrapolated to the general population.
Moreover, although the 1993 New England Journal of Medicine study purported to show that the homicide victims were killed with a gun ordinarily kept in the home, the fact is that as Kates and associates point out 71.1 percent of the victims were killed by assailants who did not live in the victimsı household using guns presumably not kept in that home.6
While Kellermann and associates began with 444 cases of homicides in the home, cases were dropped from the study for a variety of reasons, and in the end, only 316 matched pairs were used in the final analysis, representing only 71.2 percent of the original 444 homicide cases.
This reduction increased tremendously the chance for sampling bias. Analysis of why 28.8 percent of the cases were dropped would have helped ascertain if the study was compromised by the existence of such biases, but Dr. Kellermann, in an unprecedented move, refused to release his data and make it available for other researchers to analyze.
Likewise, Prof. Gary Kleck of Florida State University has written me that knowledge about what guns were kept in the home is essential, but this data in his study was never released by Dr. Kellermann: "The most likely bit of data that he would want to withhold is information as to whether the gun used in the gun homicides was kept in the home of the victim."*
As Kates and associates point out, "The validity of the NEJM 1993 studyıs conclusions depend on the control group matching the homicide cases in every way (except, of course, for the occurrence of the homicide)."6
However, in this study, the controls collected did not match the cases in many ways (i.e., for example, in the amount of substance abuse, single parent versus two parent homes, etc.) contributing to further untoward effects, and decreasing the inference that can legitimately be drawn from the data of this study. Be that as it may, "The conclusion that gun ownership is a risk factor for homicide derives from the finding of a gun in 45.4 percent of the homicide case households, but in only 35.8 percent of the control household. Whether that finding is accurate, however, depends on the truthfulness of control group interviewees in admitting the presence of a gun or guns in the home."6
The Problem with Scientific Surveys
Professor Gary Kleck has written extensively that false denial of gun ownership is a major problem in these survey studies, and yet Kellermann and associates do not admit or mention this fact.9 And this is critical. It would take only 35 of the 388 controls falsely denying gun ownership to make the control gun ownership percentage equal that of the homicide case households. As Kates and associates write, "If indeed, the controls actually had gun ownership equal to that of the homicide case households (45.4 percent), then a false denial rate of only 20.1 percent among the gun owning controls would produce the thirty-five false denials and thereby equalize ownership."6
Consider the fact that Kellermann and associatesı pilot study had a higher percent false denial rate than the 20.1 percent required to invalidate their own study, and yet, he and his associates concluded that there was no "underreporting of gun ownership by their control respondents," and their estimates, they claim were, therefore, considered not biased.4
In the Medical Sentinel, we have considered this type of bias** in response to a JAMA 1996 gun ownership survey. We reported on question #20 of that survey: "If asked by a pollster whether I owned firearms, I would be truthful? 29.6 percent disagreed/strongly disagreed."10 So according to this survey, 29.6 percent would falsely deny owning a firearm. We know that nearly one-third of respondents intentionally conceal their gun ownership because they fear further confiscation by the police as has happened in cities such as Washington, D.C., Detroit, and New York.
One must conclude on the basis of these errors that the findings of the 1993 Kellermann study are invalidated, just as those of 1986 are tainted. Nevertheless, these errors have crept into and now permeate the lay press, the electronic media, and particularly, the public health literature and the medical journals, where they remain uncorrected and are repeated time and again and perpetuated. And, because the publication of the data (and their purported conclusions) supposedly come from "reliable" sources and objective medical researchers, itıs given a lot of weight and credibility by practicing physicians, social scientists and law enforcement These errors need to be corrected to regain the loss of credibility of public health in this area of gun and violence research.
Are There Benefits of Firearms?
What we do know, thanks to the meticulous scholarship of Prof. Gary Kleck and Doctors for Integrity in Policy Research (DIPR), is that the benefits of gun ownership by law-abiding citizens have been greatly underestimated. In Point Blank: Guns and Violence in America (1991),11 myriads of scientific publications, and his latest book, Targeting Guns (1997),9 Prof. Kleck found that the defensive uses of firearms by citizens amount to 2.5 million uses per year and dwarf the offensive gun uses by criminals. Between 25-75 lives are saved by a gun for every life lost to a gun. Medical costs saved by guns in the hands of law-abiding citizens are 15 times greater than costs incurred by criminal uses of firearms. Guns also prevent injuries to good people and protect billions of dollars of property every year.5***
Incidentally, the health care costs incurred by gun shootings have been greatly exaggerated. DIPR, in an article published in the June 1995 issue of the JMAG, estimated that the actual U.S. health care costs of treating gunshot wounds is approximately $1.5 billion which amounts to 0.2 percent of annual health care expenditures. The $20-$40 billion figure, so frequently cited by the mass media, and even medical journals, is an exaggerated estimate of lifetime productivity lost where criminals are given inflated, unrealistic life productivity estimates, as if their careers were suddenly expected to blossom into that of pillars of the community12 with projected salaries equaling those of managed care CEOs. Yet, despite these major detractions, the health advocacy establishment clings to the erroneous figures and extrapolations of Dr. Kellermann and other public health researchers, and use these erroneous figures in propounding health and gun control policies, to the detriment of public policy.
To catch up with the lost ground on the gun and violence research that has been accumulating in the criminologic and sociologic body of literature in the last couple of years, we have to look not only to the data collected by Prof. Gary Kleck and Dr. Edgar Suter, but also other prominent investigators. Recent data by Prof. John R. Lott, Jr., formerly with the University of Chicago and now at Yale University, in his book More Guns, Less Crime Understanding Crime and Gun Control Laws (1998)13 has also been suppressed from dissemination in the medical journals and public health literature, except for the Medical Sentinel.1-2 In his book, Prof. Lott studied the FBIıs massive yearly crime statistics for all 3,054 U.S. counties over 18 years (1977-1994), the largest national survey on gun ownership and state police documentation in illegal gun use, and he comes to some startling conclusions:
Suicide and "Crimes of Passion"
Let me now say a word about suicide and gun availability. Both Drs. Arthur Kellermann and John H. Sloan have written about suicides and have attempted to link these fatalities to the availability of guns in articles published in The New England Journal of Medicine.14-15 In reality, the overwhelming available evidence compiled from the psychiatric literature is that untreated or poorly managed depression is the real culprit behind the high rates of suicide. The evidence is authoritative on this point as classified in the Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association and any standard psychiatric text. From the social science of criminology, in fact, we solve the seeming paradox that countries such as Japan, Hungary, and in Scandinavia which boast draconian gun control laws and low rates of firearm availability have much higher rates of suicide (2 or 3 times higher) than the U.S. In these countries where guns are not readily available, citizens simply substitute for guns other cultural or universally available methods for killing oneself, such as Hara-kiri in Japan, drowning in the Blue Danube as in Hungary, suffocation (with poisonous gases such as carbon monoxide from automobile exhausts), or simply hanging like in Denmark and Germany, or even drinking agricultural pesticides as is commonly done in Sri Lanka. And in these countries, citizens commit suicide quite effectively by these methods at higher rates than in the U.S.11,16 I believe the health advocacy establishment must consider the fact that guns and bullets are inanimate objects that do not follow Kochıs Postulates of Pathogenicity (which prove definitely and scientifically a micro-organism is responsible for a particular disease), and recognize the fact that behind every shooting there is a person pulling the trigger and who should be held accountable. The problem is more complex than just easy availability of firearms and guns and bullets as animated, virulent pathogens, needing to be stamped out by limiting gun availability, and ultimately, eradicating guns from law-abiding citizens.
And, within the context of gun availability, much has been said about the "crimes of passion" that supposedly take place impulsively, in the heat of the night or the furor of a domestic squabble. Criminologists have pointed out that homicides in this setting are the culmination of a long simmering cycle of violence. In one study of the police records in Detroit and Kansas City it was revealed, for example, that in "90 percent of domestic homicides, the police had responded at least once before during the prior two years to a disturbance," and in over 50 percent of the cases, the police had been called five or more times to that dysfunctional domicile. Surely, these are not crimes of passion consummated impulsively in the heat of the night by ordinary citizens, but the result of violence in highly dysfunctional families in the setting of repeated alcohol or illicit drug use; it is also the setting of abusive husbands who after a long history of spousal abuse finally commit murder, and increasingly, wives defending themselves against those abusive husbands, representing acts of genuine self-defense.6 Part II follows.
All references cited in this text are found at the conclusion of Part II of this essay.
* Personal communication via e-mail, 09/21/99.
** Because of the reluctance of some investigators even those funded by taxpayers to share scientific data with other researchers, the Medical Sentinel, the official, peer-reviewed journal of the Association of American Physicians and Surgeons (AAPS), established the open data policy for public review of research impacting on the formulation of public policy, http://www.haciendapub.com/press2.html.
*** This has been substantiated by a Department of Justice study in 1997 under the Clinton administration which found that up to 1.5 million citizens use firearms to protect themselves and their property yearly.