St. Louis University Public Law Review
Crime and Punishment Symposium,
vol 12, no. 2, 1993: 243.
Posted for Educational use only.
The printed edition remains canonical. For citational use please visit the local law library or obtain a back issue.
WHY WE SHOULD BAN HANDGUNS IN THE UNITED STATES
Nicholas Dixon *Copyright © 1993 by the Saint Louis University School of Law & Nicholas Dixon
Readers of this review are likely to be familiar with the controversy over whether restrictions on gun ownership are compatible with the Second Amendment's guarantee of "the right to bear arms." [1] There would be little point in discussing the complex question of the constitutionality of gun restrictions, however, unless there were good reasons for implementing them in the first place. The purpose of this paper, which will be confined to handguns, is to argue that there are good reasons for the most stringent restriction - an outright handgun ban. This paper can thus be viewed as motivating and setting the stage for the constitutional debate.
My argument for banning handguns is utilitarian: the likely good consequences of my proposal, I argue, far outweigh the possible bad consequences. My main focus will be on homicide, but I will also sometimes discuss robbery and assault in connection with handguns. Aside from my detailed discussion of existing literature and evidence, I hope to advance the debate over gun control in two main ways. First, I have gathered original data, strongly supporting my hypothesis, on the correlation between handgun ownership and handgun homicide rates in various countries. Second, I have placed the discussion of rival interpretations of the evidence in the context of an elementary discussion of the nature of confirmation of hypotheses in the social sciences, and of the burden of proof that falls on their proponents and opponents. While the points I make in this regard are indeed elementary, they have been persistently ignored by opponents of gun control. In keeping with the utilitarian nature of my argument, the majority of my paper is devoted to a discussion of empirical data. However, the theoretical key to my argument is my [Page 244] brief account of confirmation and the burden of proof in section one, to which I will refer throughout the paper.
I. INITIAL ARGUMENT
A. Introduction
In 1990 there were 23,438 homicides in the United States, 9,923 of which are known to have been committed with handguns. [2] Of the 639,271 robberies in the United States in 1990, 36.6% involved firearms, [3] while 23.1% of the 1,054,863 aggravated assaults were made with guns. [4]
THESIS There are strong reasons for believing that one of the major causes of these 9,923 murders is the extremely high rate of private ownership of handguns in the United States. Similarly, this high rate is also a major cause of the 233,973 firearms robberies and 243,673 firearms assaults. Reducing the handgun ownership rate will reduce handgun violence, and hence the overall number of violent crimes. The most effective way to achieve such a reduction is a ban on the private ownership of handguns, with exceptions narrowly confined to the armed forces, the police, private security guards, and licensed gun collectors.
A ban on the private ownership of handguns will restrict the freedom of United States citizens and require an adjustment in the way that some of them spend their leisure time. I accept that the burden of proof is on me to demonstrate that the benefits of my proposal outweigh its costs. I discharge this burden in the rest of section I and reinforce my response throughout section II. Having shifted the burden of proof to opponents of gun control, in section II I discuss responses that have been given to arguments for a handgun ban. My argument is primarily a utilitarian discussion of the beneficial consequences of a handgun ban (a reduction in the murder rate and a general decrease in violent crime, especially robbery and aggravated assault). The pleasure and additional self-defense which is alleged to result from owning and using handguns is trivial compared to the death and misery that is caused by their misuse. However, my thesis could be equally well expressed in terms of rights (the right to life, freedom from assault, and property of victims of handgun crimes). The restriction of the alleged right to bear arms is minor compared [Page 245] to the violations of the rights of the victims of handgun crimes that occur every day.
I have focused on a handgun ban primarily because handguns are the weapon of choice of violent criminals. In 1990 handguns were used in 77.2% of murders involving firearms and 49.5% of all murders in the United States. More recent figures are not available, but in 1967 96% of firearms used in robberies and 86% of those used in aggravated assaults were handguns. [5] These numbers are almost certainly attributable to their relative cheapness, their small size (and hence greater concealability), and the fact that they are easy to use. At the same time, long guns (shot guns and rifles) are used more than handguns in recreational pursuits, which, ceteris paribus, it would be desirable to allow to go unhindered. Consequently, and in view of their minimal criminal use, I see no pressing need for a ban on long guns. Because of the high percentage of violent crimes that are committed with handguns, and because they are uniquely suited to such use, a handgun ban will result in a reduction in overall rates of violent crime. [6]
Many recreational uses of handguns are compatible with a ban on private ownership. For instance, target shooting can still be enjoyed at licensed facilities. Shooters would be allowed to own or rent handguns that would be permanently stored at the shooting ranges. [7]
Licensed gun collectors would be allowed to keep handguns of recognized antique value-say fifty years old or more-on the strict condition that no ammunition be kept. Another cost of a ban would be that gun dealers would lose the profits they currently make from the sale of handguns and their ammunition. However, they would be able to recoup some of these losses by diversifying their stock of long guns, which would be unaffected by my proposal. The reduction in violent crime that would result outweighs whatever loss of profits may occur for this relatively affluent sector of our society.
James B. Jacobs has raised serious questions about handgun bans that allow for exceptions. [8] Since I advocate exceptions, I will [Page 246] consider and respond to his arguments. He points out that the widely- publicized handgun prohibitions in San Francisco and Morton Grove, Illinois, allow for extensive exceptions. In addition to the exceptions I allow, these prohibitions permit people such as prison guards, campus police, sheriffs, police chiefs, and retired law enforcement personnel to keep handguns at home and at work, and San Francisco allows business owners or employees to carry handguns on business premises. [9] Such exceptions, argues Jacobs, call into question the fairness of denying guns to other people who may feel that they have an equally good reason to own handguns. Moreover, "the exceptions reinforce the apparently widely held feeling that handguns are valuable, even necessary, for personal security." [10] Jacobs even goes so far as to suggest that we cannot even allow the police to use handguns "without letting loose tremendous pressure to recognize other self-defense and safety claims that could ultimately swallow up the prohibition." [11] As a result, such bans will lack credibility in the eyes of the public, who will freely disobey them. [12] A credible, effective handgun ban must be total and include the police, whose leadership will be crucial in changing people's attitudes towards guns. It is even more important to disarm private security guards.
If policymakers acknowledge the value of handguns for self-defense and crime prevention by allowing private security personnel to carry handguns, however, the moral coherence of handgun prohibition is seriously, perhaps fatally, weakened. Large number of private citizens remain unshakably convinced that they need handguns in their stores and in their homes to deter or thwart criminal predators. [13]
My strategy in responding to Jacobs' slippery slope argument is to show that there are relevant differences between those people whom I would allow to use handguns and the rest of the population. I consider first the police. Like Jacobs, I would welcome the day when the United States police, like the British police, do not need to routinely carry guns. I also appreciate his efforts to explore ways that police could protect themselves, carry out arrests, and deter crime without carrying guns. [14] However, because long guns will still [Page 247] be legal and widely owned, and some criminals can be expected to keep their handguns until caught even if my ban is enacted, police will have to remain armed for the foreseeable future in order to adequately enforce the law. It will take many generations, maybe centuries, before we achieve the same situation as in countries like England, which do not have a tradition of private handgun ownership. The moral legitimacy of allowing the police a deadly force which is denied to the general population is easily established by reference to the special role of the police. Since the police are entrusted with the protection of society and the prevention and deterrence of crime, it is only to be expected and indeed encouraged that they be given force superior to the rest of society. Since they are subject to extensive training and strict discipline, police are less likely to abuse handguns than the private citizen, including both criminals and law-abiding citizens who own guns for self-defense.
Private security guards would be allowed to use handguns while on duty and keep them at home, safely locked and unloaded. The justification for distinguishing between ordinary citizens and private security guards is again the nature of their work. Security guards are entrusted with the protection of large amounts of cash or other valuables, and by virtue of their uniforms and their place of work, they are visible targets for attack. Consequently, they need to be at least as well armed as potential attackers. A further distinction between ordinary citizens and private security guards is the training that the guards would undergo. Furthermore, private security companies would have to be licensed, and a condition of such licenses would be precisely that they give adequate training to their employees.
Owners of convenience stores, gas stations, and other small businesses, which are the target of an increasing number of armed robberies, can also reasonably claim a need for handguns on their premises for self-protection. However, the danger of over-zealous use of firearms in response to perceived threats of robbery, [15] and the general undesirability of adding to the pool of privately owned guns that may fall into the wrong hands, argue against this exception. Security guards are specialists in the protective use of firearms and are far less likely to abuse their weapons. It is reasonable to expect business owners who feel they need such protection to hire private security guards. These guards can transport cash takings to the bank [Page 248] and provide permanent protection at the place of business. Hiring such protection would be one of the expenses taken into account by those contemplating going into business and desiring armed protection.
B. An International Comparison
In 1988 Interpol reported the following number of handgun homicides for these countries:
Country Handgun Homicides
Population [16]
Rate per 100,000
Australia 13
13080
0.07
Canada 8
25,857,000 (1987)
0.031
Great Britain 7
57,376,000 (1990)
0.012
Israel 25
4,614,000 (1990)
0.542
Sweden 19
8,332,000 (1984)
0.228
Switzerland 53
6,473,000 (1985)
0.819
United States 8,915
250,410,000 (1990)
3.560
It was this astounding disparity between the United States and other developed countries which first drew my attention to the issue of hand gun control.
My contention is that a major cause of this disparity is the much higher rate of handgun ownership among private citizens in the United States compared to other countries. More generally, I argue that any country's handgun ownership rate is a major determinant of its handgun homicide rate. The following table is based on information [Page 249] from government agencies, including police departments, in the respective countries. Any systematic bias, which may result from a government agency's desire to minimize or exaggerate gun ownership levels, can reasonably be assumed to apply equally to all of the countries studied. Since my interest is in comparative ownership rates rather than the absolute numbers, any such bias is irrelevant. The numbers refer to estimates of the total number of handguns owned by civilians in each country, both legally and illegally.
Country Handgun
Handguns per 100,000
Handgun Homicides per 100,000
United States 56,833,000 [17]
22,696
3.56
Israel 171,448 [18]
3,716
0.542
Sweden 308,261 [19]
3,700
0.228
Canada 595,000 [20]
2,301
0.031
Australia 263,900 [21]
1,596
0.07
Great Britain 480,000 [22]
837
0.012
[Page 250]
(The Swiss government was unable to provide any handgun ownership estimates.)
My handgun ownership estimates (except the number for Great Britain) are the result of independent inquiries to government agencies in each country. The close coincidence between the rank ordering of handgun ownership and handgun homicide rates in these six diverse countries is most plausibly explained by the causal connection I assert. The multiplicity of causes of handgun homicide which opponents of handgun control are eager to assert, and which I accept, make a perfect correlation most unlikely. The one anomaly is the relative position of Australia and Canada; but the actual handgun homicide numbers (13 and 8, respectively) are so small as to make the difference in homicide rate of little importance.
Of more interest than the rank ordering of individual nations, [Page 251] which is more sensitive to "interference" from other causal factors, is the emergence of three clear categories in which handgun ownership and handgun homicide rates coincide: low (Canada, Australia, and Great Britain), moderate (Israel and Sweden), and high (United States). The most significant fact of all is the vast disparity between the United States and all the comparison countries in both the handgun ownership and handgun homicide rates. I conclude that a dramatic reduction in the handgun ownership rate in this country would substantially reduce handgun homicide rates.
I am assuming that the number of handguns in a country depends on (1) the permissiveness of its handgun laws, and (2) the demand for handguns. Handgun laws in the United States are far more permissive than in any of the comparison countries. [23] Since the law is much more easily controlled than the people's wishes, by far the easiest way to reduce handgun ownership is to pass more restrictive laws. My proposal, then, is that the best way to reduce handgun homicides is to pass maximally restrictive laws - a handgun ban.
Two interesting points concerning the demand for handguns are worth noting. First, it is probable that, doubtless due in part to the long history of private gun ownership in this country, there is more demand for them in the United States than in the other countries. [24] In order to achieve the same levels of gun ownership in the United States as in other countries, therefore, it is likely that even more restrictive handgun laws will be required. Second, a reduction in the number of handguns in this country (by means of a handgun ban) can reasonably be expected to result in a reduction in demand, which will in turn cause a further reduction in ownership levels. This result is because a major reason for handgun ownership at present is to defend oneself against the huge number of people who already have handguns. (See infra section II.E for a discussion of the defensive efficacy of handguns.) I propose stemming this spiral of gun ownership at its source rather than simply acquiescing in the unlimited proliferation of handguns.
Two important clarifications need to be made at this point. First, it is not being claimed that the high rate of gun ownership in the United States is the only, or even the main, cause of its exceptionally [Page 252] high handgun homicide rate. What is being claimed is that its handgun ownership rate is one of the causes. Furthermore, it is the easiest to control of all of the probable causes. Consequently, reducing ownership of handguns is the most realistic way to start reducing murder and handgun-related crime in the United States. Second, I am fully cognizant of the error of assuming that a correlation implies a causal connection. In order to avoid this error, anyone who posits a causal connection based on a correlation must do at least two things. One must first show that there are no other variables which correlate better with the effect, and which would account for the effect better than, or in place of, the posited cause. Take, for instance, the view that AIDS was a punishment for homosexuality, which, in the early 1980s, did seem to correlate very closely with the syndrome of diseases. As heterosexual AIDS cases emerged, the "punishment" hypothesis lost credibility. What finally destroyed the credibility of that hypothesis was the discovery of a 100% correlation between AIDS symptoms and the presence of HIV, along with the emergence of an increasing number of heterosexuals with AIDS. The second requirement is the provision of a probable theoretical explanation of how the causation occurred. The second requirement is also illustrated by the case of AIDS. What made the HIV hypothesis increasingly convincing was the development of a detailed biochemical explanation of exactly how the virus attacks the immune system and leads to the symptoms of AIDS. Both the 100% correlation and the detailed theoretical account make the HIV hypothesis practically certain. In the case of causal hypotheses in the social sciences, where the web of causation is much more complex and causes much harder to distinguish, a more modest correlation and a less rigorous theoretical explanation are sufficient to establish the plausibility of a causal connection.
Since I do not claim that handguns are the only cause of murder, I do not need to rule out the existence of other causes. Consequently, to try to refute my position by pointing out these other causes is to commit a straw man fallacy. All I need to show is that there is no other cause that correlates so well with handgun murder as to rule out my own causal hypothesis. I undertake this task in the next subsection and on various occasions throughout this paper. While the evidence does indeed suggest a prima facie case for several other causal factors, none of them is nearly strong enough to be considered as the only cause, and hence, to disprove my hypothesis. As for a theoretical explanation of why high rates of handgun ownership correlate with high rates of handgun related murder, one need not go beyond common sense. Assuming human nature to be relatively similar in different developed democratic countries (i.e. [Page 253] those represented in the Interpol statistics quoted above), one would expect people to be subject to roughly similar amounts of stress, provocation, jealousy, anger, desperation, resentment of other people's affluence, and whatever other factors are liable to lead some people to violence. If one of these nations has a vastly higher rate of private ownership of handguns, one would expect that the similar provocations to violence would spill over into handgun murder far more often than in the other nations. This low-level theoretical explanation is sufficient to show that my handgun hypothesis is more than an accidental coincidence and, unlike the "AIDS as punishment" hypothesis, is not based on prejudice and superstition.
C. The Burden of Proof
I have presented evidence of a striking correlation between the rate of private ownership of handguns and the rate of handgun murder in six different countries. I have given a theoretical account of why this correlation is a causal one. Throughout the rest of my paper, I will show that none of the other alleged causes of the high homicide rate in the United States comes even close to disproving my hypothesis that the high ownership rate of handguns is one of its significant causes. By the end of my paper I will have amply met the burden of proof that I set for myself above - namely, showing that a ban on handguns will reduce our homicide and violent crime rates, thus justifying the minor restriction on liberty that it will involve.
I now wish to go on the offensive and suggest that I have already written enough to issue a different burden of proof challenge to opponents of a ban on handguns, in light of the strong prima facie case I have made for my causal hypothesis. In the case of my comparative international homicide statistics, and of the other statistics that I will adduce throughout my paper, it is incumbent on them to produce an alternative causal account proving that the United States' high handgun murder rate is caused by factors unrelated to its high rate of handgun ownership. They must specify what these causes are, quantify their relative presence in the United States as compared to the countries with lower homicide rates, demonstrate that variations in these factors correlate with variations in the murder rate, and provide a plausible theory explaining the causal mechanisms at work. I contend that they have utterly failed to even approach a satisfactory response to this burden of proof.
A strategy commonly used by opponents of gun control is to construct thought experiments in which the evidence gathered by advocates of gun control can be attributed to factors other than guns. Except in the case of deductively valid arguments, it will always be [Page 254] possible to describe scenarios in which the premises are true and the conclusion false. When dealing with the highly complex webs of causation that are at work in crime and violence, the best we can hope for is arguments that offer reasonably high degrees of inductive support for their conclusions. The counter examples offered by opponents of gun control certainly show that the arguments in favor of gun control are not deductively valid; but no one has ever claimed that they are. [25] Scenarios unsupported by statistical evidence and theoretical explanations of the causal mechanisms do not even address the burden of proof I have asserted. The only way opponents can disprove my hypothesis is to provide such evidence and explanations and show that they are more probable than the causal explanations I provide.
D. Comparison with Capital Punishment
In order to illustrate the burden of proof which I place on opponents of a handgun ban, I turn briefly to the criteria by which we decide whether the effectiveness of capital punishment (CP) as a deterrent to murder has been established. The available evidence indicates that there is no significant correlation between a jurisdiction's murder rate and whether or not it has the death penalty. This lack of correlation is strong prima facie evidence against the main utilitarian argument for capital punishment: that it will reduce the murder rate.
Proponents of CP have responded that its superior deterrent effect is "masked" by a complex of other factors. For instance, the similar murder rates in states with and without CP is explained by CP's "general deterrent" effect, which allegedly operates even in those states that do not have CP. [26] Sociopolitical differences between these states may also serve to mask CP's superior deterrence. Without CP, some states would likely have a far greater than average murder rate that is currently kept in check precisely because of the operation of CP. Prima facie support for these contentions is provided by considerations such as the "Preference Argument," according to which it is clear that, given the choice, nearly everyone would [Page 255] choose life imprisonment rather than execution. [27] Together, these considerations are used to argue that CP is indeed the best deterrent to murder, even though various other factors combine to obscure this deterrent effect.
This brief discussion of the deterrent effect of CP illustrates how the burden of proof principles which I outline above apply to a concrete situation that is in some ways analogous to the handgun control debate. First, given the strong empirical evidence that CP is not a superior deterrent, the burden of proof is on those who believe that it is. They need to produce plausible explanations of why CP is the best deterrent, even though the evidence fails to bear this out. It is fair to say that supporters of CP have been reasonably successful in suggesting causal hypotheses that account for the lack of empirical support for their belief. However, they have fallen short of showing that these causal hypotheses as in fact correct. In other words, showing that the evidence is compatible with CP's superiority as a deterrent fails to meet the burden of proving that it is in fact a superior deterrent. This, I contend, is the situation in which those who deny my causal hypothesis concerning the prevalence of handguns and the handgun murder rate find themselves. Opponents have shown that it is possible that my hypothesis is false (which I have never denied, since it is not claimed to be a deductively valid argument); but they have failed to give any good reason to believe that it is in fact false.
Second, if, on the other hand, there were a substantially lower murder rate in states and countries that have CP, the burden of proof would shift. It would then be encumbent on opponents of CP to explain why, appearances to the contrary, CP is not a better deterrent than other punishments for murder. It seems very likely that many of those who currently oppose CP on utilitarian grounds would abandon their opposition in the face of such hypothetical evidence. The evidentiary situation with regard to the connection between the ownership of firearms and the murder rate is, on the surface, more clear-cut than in the case of CP. When comparing the United States with most other developed countries, there is an astounding disparity with regard to both the handgun homicide rate and the handgun ownership rate. By parity of reasoning, then, it is fair to insist that opponents of a handgun ban abandon their view unless they can explain why their view is correct in spite of the evidence. [Page 256]
II. RESPONSES TO THE BURDEN OF PROOF CHALLENGE
To be fair to opponents of gun control, there is a substantial literature addressing the burden of proof challenge that I have issued. Defenders of private handgun ownership have written extensively on why, appearances notwithstanding, a ban on handguns will not reduce the homicide rate in the United States. This section will be devoted to the analysis of such defenses.
A. Comparisons With Other Countries
Since comparisons with the far lower murder rates in countries that have stricter handgun control were the main impetus for gun control, it is wise to start with this issue. The overall strategy of handgun supporters is to argue that the higher murder rate in the United States compared to other developed countries is attributable to factors other than the higher prevalence of handguns in the United States. [28] For example, Don Kates, one of the most prolific and articulate opponents of banning handguns, argues: "The determinants of violence are ... fundamental economic, sociocultural, and institutional differences .... Since gun laws, by definition, do not focus on these kinds of fundamental determinants, their potential benefits can be no more than marginal." [29] Disappointingly, neither Kates nor any of the other contributors to his volume give any analysis of what exactly these "deeper" causes are and how one might hope to remedy them. Such vague hypotheses fail to meet the evidentiary burden of proof that I have placed on opponents of gun control. [30]
Other opponents of gun control try to show that international comparisons actually weaken the case for gun control. Though I could not obtain any handgun ownership statistics from the Swiss government, it is often claimed that the rate of gun ownership in Switzerland is higher than that in the United States. [31] This is alleged [Page 257] to disprove any causal connection between firearm ownership and homicide rates. However, there are two crucial differences between this country and the United States. First, the guns owned in Switzerland are primarily long guns. [32] Long guns are not the issue in this paper, which advocates a ban on handguns only. Second, all male citizens in Switzerland are required to retain the gun that they were given during their military service. [33] The context of their gun ownership is, then, mandatory service in a citizens' militia, with its attendant training and discipline, which bears no comparison with the minimally controlled private handgun ownership in the United States. In fact, this comparison was most ill- advised on the part of handgun supporters. The handgun homicide rate in Switzerland, though less than that in the United States, is almost four times higher than that in Sweden and is on average over ten times higher than that in other countries with restrictive handgun laws (Australia, Canada, and Britain). The factors to which opponents of gun control appeal in order to explain the high rate of handgun homicide in the United States-e.g. extensive poverty, high unemployment, a minimal welfare system, and racial tension-cannot plausibly be asserted of Switzerland. [34] The United States' alleged high rate of firearms ownership remains the most plausible explanation of its comparatively high handgun homicide rate.
More importantly, it is not necessary for me to respond to these and other attempts to discredit my international comparison by reference to causes of crime that are unrelated to gun laws. I have already made clear that I do not deny that factors other than the prevalence of handguns may influence the rate of violent crime. It should be no surprise that these factors prevent a uniform correspondence in all countries between levels of gun ownership and violent crime. Advocating a ban on handguns is perfectly compatible with recognizing that a concerted attack on unemployment, homelessness, huge disparities in wealth and real opportunity, racial inequality, and other sources of injustice are of much greater importance in the attempt to reduce homicide and violence.
My central thesis is that a major cause of the high handgun homicide rate in the United States is its huge arsenal of privately owned guns, and a handgun ban would be the best way to reduce this arsenal. I have presented striking empirical data to support my causal hypothesis. The burden of proof that I have charged to opponents [Page 258] of gun control is certainly not met by pointing out the existence of some causes of murder and violent crime that are not addressed by gun control. In view of the fact that the deeper socio- economic causes of violent crime are very difficult to control, we need to address other causes that are amenable to control. The availability of firearms is one such factor that can be controlled by legislation. It is ironic that opponents of a handgun ban point out these deeper, more institutional causes of violence in the United States, as if they somehow show the pointlessness of remedial measures. On the contrary, they only serve to underline the need for strict handgun control measures.
Scepticism as to the value of international comparisons concerning gun control and gun related crime is even less plausible in light of a study done in 1988. [35] In order to isolate the key variable-the impact of gun control on violent crime in general and on firearm violence in particular-the authors studied two cities that are very similar in most other respects: Seattle and Vancouver, Canada. [36] The two cities have a similar population, geography, climate, level of schooling, unemployment rate, median annual household income, and cultural values. [37] Of particular interest, however, is the great similarity in their overall crime statistics. Vancouver had a very slightly higher burglary rate, and in other types of crime, Seattle had a slightly higher relative risk: robbery (1.09:1), simple assault (1.18:1), and aggravated assault (1.16:1). [38] With regard to the weapons used in aggravated assaults, both cities reported almost identical rates of assaults with knives, other dangerous weapons, and hands and feet. [39] These similarities are in precisely the same factors to which gun control opponents usually appeal in order to account for the higher rate of gun violence in the United States. At this point the similarities in crime patterns end. In the period studied, Seattle had 11.3 homicides per 100,000 person-years, whereas Vancouver had 6.9 per 100,000 person-years. [40] Consequently, the relative risk of being [Page 259] murdered in Seattle as compared to Vancouver was 1.63:1. [41] The relative risk of homicide excluding those committed with firearms was very similar (1.08:1), but the risk of being murdered with a firearm in Seattle as compared to Vancouver was 4.8:1. [42] Eighty-five percent of the firearms homicides in both cities were committed with handguns. [43] It will be difficult to deny that the almost fivefold difference in the frequency of homicides committed with firearms is responsible for the substantially higher homicide rate in Seattle. [44]
One marked difference between the two cities is that Vancouver, like all of Canada, has significantly stricter gun control laws. [45] The most important difference is that Vancouver does not allow concealed weapons and grants handgun permits for sporting and collecting purposes only. [46] Handguns may be transported by car only if they are stored in the trunk in a locked box. [47] In Seattle, concealed weapons are allowed with a permit. [48] This has resulted in a disparity in the rates of gun ownership in the two cities. In the 1984-88 period, the total number of handgun permits issued in Vancouver was 4137. [49] In the same time span, Seattle issued 15,289 concealed-weapons permits; in addition, no permit at all was needed for handguns kept at home. [50] An independent measure of gun ownership is provided by "Cook's gun prevalence index," which is based on surveys and the number of suicides, assaults, and homicides involving firearms in forty-nine cities in the United States. The index assigns a 41% gun ownership rate to Seattle, and only 12% to Vancouver. [51] To summarize, we have two cities which closely resemble each other in terms of sociology, population, economics, culture, and overall crime patterns, including nonhomicidal violent crime. However, there is a noticeable disparity in their rates of homicide and a hug difference in their rates of gun-related homicide. The city with the [Page 260] lower homicide rates has far stricter gun control laws (especially for handguns, which were responsible for 85% of the firearms- related murders in both cities), and, not surprisingly, a far lower rate of gun ownership. The burden is on opponents of gun control to show why this study does not demonstrate the link between rates of gun ownership and homicide rates.
It is to the credit of opponents of gun control that they have taken on the challenge presented by the Seattle-Vancouver study. James Wright has criticized it on two main grounds. First, it fails to prove that guns are more widely available in Seattle than in Vancouver. Second, the difference in murder rate is attributable to racial factors, not to differences in gun laws. [52] To establish the first criticism, Wright points out that the study's reason for believing that there are more guns in Seattle than in Vancouver is based on the number of gun permits handed out in the two cities. He discounts these numbers on the ground that the cities have different permit regulations. [53] However, the very difference between permit regulations indicates that the figures vastly under-estimate the number of handguns in Seattle, since Seattle, unlike Vancouver, requires neither permits nor registration for handguns kept at home.
Wright's second main line of objection is based on a comparison of the homicide rates of Seattle and Vancouver by ethnic groups. While they have similar percentages of white residents (Seatle:79.2%, and Vancouver:75.6%), the makeup of their non-white populations is very different. Vancouver is dominated by Asians (22.1%), whereas Seattle has a higher percentage of blacks (9.5-0.3%) and Hispanics (2.6-0.5%). [54] Revealingly, despite its overall substantially lower murder rate, Vancouver's annual homicide rate for its white non-hispanic population is slightly higher than Seattle's (6.4 to 6.2 per 100,000.) [55] If the alleged difference in gun ownership were the cause of the difference in overall murder rate, one would expect this effect to be reflected among all racial groups, whites included. Since the difference appears only among racial minorities, the evidence indicates racial differences, not differences in gun laws, as the cause of the difference in murder rates. [Page 261]
However, the existence of these racial variations does not invalidate the Seattle-Vancouver study. Its authors are perfectly well aware of these differences and suggest that the usual socio-economic disadvantages of these groups are part of the cause. [56] These disadvantages make members of racial minorities more likely to commit murders, and the vast majority of murders are committed by someone of the same ethnic group as the victim. [57] Their own statistics show that, with the exception of Asians in Vancouver, all non-white groups in both cities suffer from a higher murder rate than whites. [58] No one in the gun control movement ever claims that the availability of guns is the only cause of murder. The point is that it is one of the causes that exacerbates the other causes such as socio-economic deprivation. In support of this hypothesis is the fact that the non- white/white disparity in murder rate is much more pronounced in Seattle (5.78:1) than in Vancouver (3.63:1). [59] The difference in the white/non-white disparity is plausibly explained by the far greater prevalence of handguns in Seattle. It is among those elements of the population who, by virtue of disadvantages linked to race (discrimination, lack of economic opportunity, poverty, unemployment, and so forth), are more likely to kill that we should expect the homicide-increasing influence of handguns to be most pronounced. It is, as it were, "the straw that breaks the camel's back" in the case of violence-prone sectors of society. [60] It is less surprising that there is no perceptible influence of guns on white people who, statistically, are less prone to be murdered in the first place. [61] Moreover, Seattle and Vancouver are anomalous in this regard. Nationwide, "the homicide rate among whites alone is almost three times higher in the U.S. than in Canada." [62] The "racial differences" hypothesis is of no use in explaining this disparity.
It is worth noting that the burden of proof issue is again relevant with regard to the Seattle-Vancouver study. The study provides [Page 262] a documented correlation between gun availability and homicide rates, which is, furthermore, supported by a theory explaining why the easy availability of guns is a major cause of a high murder rate. The critics assert that the real cause of the difference in homicide rate may be unrelated to guns, and related instead to differences between the situation of non-white populations in the two cities. They do not specify exactly what these differences are, nor do they quantify these differences and provide empirical evidence to support their numbers. In any event, then, they have failed to provide serious reasons to doubt that the study provides strong support for banning handguns.
B. Inter-State and Inter-City Comparisons
Further support for gun control would be provided by statistics demonstrating higher gun-related homicides and other violent crime rates in those states with more permissive gun laws. However, it is widely agreed in the anti-control literature that the evidence fails to provide any such support. [63] In response, advocates of gun control suggest that the genuine deterrent effect of a state's strict gun control laws may be "masked" by the easy availability of guns in other states (especially adjacent ones) that have permissive gun laws, or none at all. Such guns can be easily "imported" from one state to another. What is called for, then, is a national gun control policy, not the abandonment of all gun control. Indeed, a plausible causal story would see it as no coincidence that states with strict gun control have high crime rates, since it is precisely in response to these high crime rates that gun control is more likely to be introduced in the first place. It is perverse to assume the contrary-namely, that gun control is itself the cause of increased crime.
However plausible such responses on behalf of gun control may be, Kates and Benenson dismiss them as the "adjacent state" excuse. [64] They argue that these responses do indeed sketch possible scenarios, but they fail to meet the burden of proof that falls on advocates of gun control. Given that they are proposing a restriction [Page 263] on liberty, and that violators of a handgun ban would be subject to severe punishment, it is encumbent on advocates of such a ban to prove that these burdens would be outweighed by even more substantial benefits. I accept this burden of proof as applicable to those who would restrict ownership of handguns. However, the starting point of this paper is precisely that handgun ban advocates have met this burden by producing statistics showing an overwhelming disparity in the murder rate between the United States and developed countries with far stricter gun controls. This, in combination with the powerful Seattle-Vancouver study, has, I argue, turned the burden of proof over to opponents of gun control. They must now respond to the immensely plausible argument that posits the easy availability of handguns as a major cause of violent crime.
To make the burden more onerous, and to further respond to the "adjacent state excuse" argument given by Kates and Benenson, I refer to the results of a May 1976 study done by the Bureau of Alcohol, Tobacco and Firearms. [65] The study traced the origin of all handguns seized by police in sixteen cities from July to December 1973, in order to determine the last retail dealer who sold the gun. Seizures would result from such causes as arrests for crimes in which the gun was used, other arrests resulting in the discovery of illegal gun possession, or from sting operations in which undercover police agents purchased guns from illegal dealers. The results strongly confirm the hypothesis that there is indeed a causal connection, which is hidden by an influx of out of state guns, between strict gun controls and the use of guns in crime. "The strength of firearm regulations or enforcement of those regulations in the differing project cities is directly proportional to the percentage of crime guns that were purchased in another State." [66]
The most striking results come from "Phase 1" of the study comparing Atlanta, Detroit, New York City, and New Orleans. Atlanta had no requirements to register and obtain licenses for handguns, and 81% of all the guns seized by the police were from Georgia. [67] (The only restriction in Atlanta was a ban on some "Saturday Night Specials.") The New Orleans Municipal Code, unlike Atlanta, does require handgun permits; but Louisiana does not, and 63% of the handguns seized in that city were from Louisiana. [68] In sharp contrast, both Detroit and New York City, and their states of Michigan and New York, have strict gun controls. Only 8% of the guns [Page 264] seized in Detroit were from Michigan, [69] and only 4% of the New York City guns were from New York State. [70] The conclusion that criminals in these two cities were deterred by strong gun controls, and they were forced to go out of state to get their handguns, is hard to resist. While gun controls operate only on the state or local level, they will constantly be undercut by the possibility of gun importation from states with more permissive gun laws. In order to prevent this, and to maximize the evident benefits of strict gun control laws, a nationwide gun control law is the obvious policy. [71]
There is always the danger, of course, that strict national gun controls will increase the number of handguns illegally imported into the United States. This is indeed a disturbing possibility, but it hardly justifies the abandonment of any attempt to control guns in the United States. An effective reductio ad absurdum of this argument can be made by considering a parallel argument which could be given against any attempt to control, for example, the manufacture of LSD, amphetamines, or barbituates in this country. After all, it seems likely that any success we achieve would be accompanied by an increased demand for and supply of illegally imported drugs. This possibility would not be seriously entertained as a reason for giving up the fight against dangerous drugs made in the United States. What is called for is an assault on both domestic and imported drugs. In exactly the same way, attempts to control the handguns made and sold in this country in no way preclude a crackdown on illegal imports of firearms, by means of devoting more resources to detection, and of heavier penalties for those apprehended. [72]
C. Handguns, Criminals, and Law-Abiding Citizens
A more troubling argument, however, concerns the different impact that a handgun ban is likely to have on criminal and law-abiding citizens. It seems plausible to suggest that law-abiding citizens are more likely than criminals to voluntarily comply with gun [Page 265] control laws, including outright bans on handguns, which will require owners to turn in their guns to the police. Criminal gunowners have already committed felonies or intend to use their guns for the commission of felonies in the future, and the fact that they are committing a further felony by keeping their guns will not force compliance with handgun control laws. Indeed, the penalty for possessing an illegal gun is likely to be minimal compared to penalties that criminals face should they be apprehended for the more serious crimes that they intend to commit with the help of their guns. The very fact that they have bought guns for this purpose indicates that they are not deterred by the heavier penalties for the felonies that they plan to commit. Furthermore, drying up legal access to handguns will effectively prevent normally law-abiding citizens from becoming new handgun owners. In contrast, criminals are likely to have access to illegal black market guns and will not hesitate to avail themselves of it. The very success of a handgun ban in reducing the existing "pool" of handguns will thus result in a higher percentage of them being owned by criminals. [73] The likely result of gun control, then, especially an outright ban on handguns, is to disarm the general population, while criminals remain just as heavily armed as they are today. [74] No matter how effective a gun ban is in reducing the number of handguns in circulation, "the number of potential misusers is so small that the number of firearms legally or illegally available to its members will always be ample for their needs, regardless of how restrictive gun laws are or how strenuously they are enforced." [75] In the light of these plausible projections, some people who oppose a handgun ban do support measures which are targeted at precisely those criminal elements who will be most resistant to bans on guns and are most likely to abuse their guns. [76] Kleck, for instance, supports a ban on gun ownership for those with prior criminal convictions, which would leave the law-abiding gun owner undisturbed. [77]
Underlying the effort to target gun control at those who have prior convictions is the belief that it is these people who are most likely to misuse firearms in the future, especially in the case of homicide. However, in the case of homicides, this belief is vigorously [Page 266] challenged by advocates of gun control. Murder, the argument goes, is not confined to the ranks of those with criminal records. It is an act of terrible violence of which we are all capable if sufficiently provoked. Only 21% of murders occur during the commission of another felony. [78] In at least 48.8% of 1990 homicides, the victim was either a relative or an acquaintance of the murderer. [79] In 1990, 34.5% of all murders resulted from domestic or other kinds of argument. [80] Since we are all capable of heated arguments, we are all, in the wrong circumstances, capable of losing control and killing our opponent.
There, but for the grace of God, we all go. Given the ease with which homicide can be committed with a handgun as opposed to other more primitive methods (e.g. clubs or knives), the ease of availability of handguns may well be the factor which transforms a heated argument into a lethal attack. The simple option of running away-which is very seldom mentioned in the anti-gun control literature-will be available far more often in the case of these other kinds of attacks than in the case of a handgun attack. Gun control measures that are targeted solely at those with criminal records fail to protect us from the most likely source of handgun murder: ordinary citizens.
Such arguments are dismissed by Kleck and Bordua as "the myth of the noncriminal killer." [81] The fact that in most homicides the victim is known to the murderer does not prove that murder is in general a temporary aberration on the part of otherwise peaceful citizens. It merely reflects the fact that one is most likely to argue and fight with people one knows. [82] Criminals are just as likely to have relatives, friends, and acquaintances as law-abiding citizens. Evidence indicates that a disproportionate number of murders are committed by people with an arrest record. The number of convicted murderers with prior felony convictions is less impressive, but Kleck and Bordua's "conservative estimate" of 25% is still far more than one would expect given the presumably smaller percentage of the population with convictions. [83] "The myth of the noncriminal killer" is further weakened by a study in Kansas City of family homicides, a category one would expect to be unconnected with other criminal [Page 267] activity. In 90% of these cases, the police had been called to the same address at least once within the previous five years in response to domestic quarrels, and in 50% of the cases, police had been called in five or more times. [84] These statistics appear to indicate that gun control measures geared specifically at those with criminal records are necessary. Across the board bans would have at best a minimal impact on reducing the homicide rate, while restricting the freedom of huge numbers of peaceful gun owners.
How are we to evaluate these arguments? To begin with, the Kansas City study indicates a connection between family homicide and prior brushes with the law short of convictions. However, to deny someone access to handguns on the ground that she has been arrested (but not convicted), or that the police visited her house because of a domestic disturbance, may be unconstitutional in the light of the "equal protection" clause of the Fourteenth Amendment. A selective ban will have to focus on the 25% of convicted handgun murderers with prior felony records. This means, then, that the other 75% [85] will have the same access to handguns as they do today. Only a comprehensive ban would inhibit the access to handguns of this majority of handgun murderers. And, of course, since the ban will be comprehensive, it will include the 25% of potential handgun murderers who have been convicted for felonies. Thus my proposal includes the minor benefits of narrowly targeted gun control measures.
Another reason why a general ban is preferable to a targeted restriction is that, by virtue of reducing the overall "pool" of guns, it will reduce the real number of guns in the hands of criminals, even if it does increase the percentage of gun owners who are felons. The illegal means by which criminals would have to obtain guns-for instance buying them from unlicensed pawnbrokers, illegal transfers, buying them from friends who originally bought them legally, and outright theft-are all dependent on the presence of a substantial supply of legally purchased handguns on the market. My proposal would shrink this supply, and hence make it increasingly difficult for criminals to obtain handguns. It would also help to keep guns out of the hands of lawbreakers who have so far eluded conviction, and would hence qualify for gun ownership under a "targeted" ban. The "cost" of my proposal is that it does restrict many gun owners who never would have used their guns to commit homicide or any other crime. However, this price is more than justified by its far greater effectiveness than felons-only bans in reducing the number of murders, [Page 268] as it gradually and over the years reduces the number of handguns in circulation in the United States and chips away at the "gun culture" that encourages their use.
At this point it will be useful to delineate the differences between my view and other positions on gun control. I agree with all those groups that advocate stricter enforcement of laws against firearms crimes, and "sentence enhancement" to further punish firearms felons. [86] The fact that I advocate any gun control is enough to distinguish me from extreme opponents of handgun ownership restrictions. More importantly, I differ from both moderate opponents of gun control such as Kleck, and organizations such as Handgun Control, Inc., who are united in supporting handgun restrictions targeted at felons, such as the handgun control component of the Crime Bill debated by the Senate in July 1991. [87] I welcome such modest proposals as a step in the right direction. However, for the reasons explained above, I believe that only a general ban on handguns (with strictly limited exceptions) will effectively reduce handgun violence.
D. Substitution of Other Weapons for Handguns
Let us suppose for the sake of argument that a ban on handguns would indeed be effective in reducing the number of handgun murders and violent crime. In the pro-gun literature it is widely denied that this would result in an overall reduction in murder and violence, for the simple reason that would- be criminals will substitute other weapons for handguns. Let us first consider the effect of such substitution on robberies.
Let us suppose that robbers turn to knives, clubs, other instruments, and their hands and feet to threaten and perhaps injure their victims. This is exactly what gun control advocates want, since these weapons are far less lethal than handguns. [88] While it is true [Page 269] that stabbings and beatings are horribly lethal in their own right, a crucial difference is that running away will at least sometimes be an option for the victim, whereas this tactic will be of little use in the face of a loaded gun. A reduction in robberies and in their degree of violence is a likely result of such a substitution.
Hardy and Kates deny this conclusion. The destructive power of handguns makes them particularly effective in intimidating robbery victims into handing over their belongings or cash. As a result, robbers with guns only use their weapons to injure their victims in 17% of robberies, and can usually rely on threats alone. In contrast, robbers with knives injure 32% of their victims, and those using clubs and other weapons injure fifty-three percent of their victims. [89] In the case of blunt instruments, a blow is sometimes used to initiate the crime. Another likely consequence of the substitution of non- firearm weapons is, ironically, that there will be more robberies and hence more attendant violence. This follows for the simple reason that handgun robberies are the most efficient. In order to reap the same profits using other weapons, several robberies will be needed. [90] If gun control leads to an increase in number of robberies and injuries, one of its main purposes is clearly defeated.
However, it is strange logic indeed that would welcome the use of more lethal weapons on the ground that they will reduce the incidence and violence of robberies. This view leads to absurd consequences. If the reduction of the number of robberies is regarded as the summum bonum, then we should distribute not just handguns but assault-style automatic weapons to criminals, who will be able to use them to become rich by means of rare, but highly efficient and relatively bloodless heists.
Opponents of a handgun ban are on firmer ground when they discuss the danger that robbers, assaulters, and other criminals will "upgrade" to long guns in the event of a ban on handguns. According to Kates and Benenson, "at a minimum, a shot fired from a long gun is four times as likely to kill as one fired from a handgun." [91] [Page 270]
Widespread substitution of long guns for handguns in the commission of crimes would dramatically increase the number of homicides and violent crimes. They calculate that if only 30% of those who attempt homicide were to switch from handguns to long guns, while the other 70% "downgrade" to knives, there would still be a "substantial increase" in homicide. If the ratio were instead 50:50, the number of homicides would double, even if none of those who used knives succeeded in killing their victims. [92] Kleck asserts that an even higher substitution rate is likely. He quotes a survey by Wright and Rossi, in which prisoners who had committed several crimes with guns were asked whether they would carry a sawed- off shotgun (which would be much easier to conceal than a regular shotgun) if they were denied access to handguns. Seventy-two percent said they would, and Kleck feels justified in concluding that such a rate of long gun substitution would in fact occur. [93]
One has to doubt the reliability of the statements of prisoners as to what firearms they would carry in certain circumstances. Macho bragging and outright lying are very likely in such situations, and relegate Kleck's projections to the status of unsupported conjecture. In view of the fact that such a small percentage of the actual murders in the United States in 1990 were committed with long guns, [94] the burden on Kleck to prove his hypothetical speculation is even heavier. As for Kates and Benenson, their projections are based on the unsupported assertion that the 70% of handgun killers who do not turn to long guns would instead use knives, the most lethal weapon other than firearms. It is more probable that at least some potential murderers would turn to less lethal weapons or their bare hands, and that some would be deterred from assaults altogether. Since Kates and Benenson ignore these probable scenarios, and since their substitution predictions are in any case purely speculative, it is safe to conclude that their estimate of the increase in the homicide rate in the event of a handgun-only ban is inflated. The conjectures offered in support of the substitution hypothesis are inadequate and fail to meet the burden of proof encumbent on opponents of my proposal.
Another reason to doubt that long guns would be used in great numbers to replace handguns in robberies, assaults, and homicides is that long guns are obviously much more difficult to conceal. A potential mugger roaming the streets wielding a long gun will [Page 271] cause everyone in sight to flee, and is likely to be quickly arrested when alarmed people call the police. Similarly, a bank robber carrying a long gun will be immediately detected by security guards, alarm systems will be triggered, and the chances of a successful robbery greatly diminished. Handguns are obviously much more convenient for the commission of such crimes. Kates and Benenson point out that most homicides occur in the home, where concealability is "irrelevant." [95] However, concealability would seem to be an important factor even in the home. Since the victim may well be unaware that the killer is carrying a concealed weapon, the "surprise factor" which is peculiar to handguns can still apply even in the home. In contrast, people can hardly be unaware that the person they are with is carrying a shotgun or rifle. Moreover, in any argument or domestic quarrel, regardless of whether the potential victim knows that the assaulter is carrying a handgun, the ease of pulling out the gun and shooting makes such arguments more likely to spill over into murder. In contrast, by the time the assaulter has gone into another room to retrieve their long gun and loaded it, the potential victim has crucial seconds in which to escape.
Another reason that the concealability of handguns is not a good reason for a handgun-only ban is proposed by Hardy and Kates in their discussion of the impact of handgun control on robberies. They point out that "[t]he difference between a long gun and a handgun is ten minutes and a hacksaw." [96] Even robberies, then, would not be diminished by a handgun ban. However, this contention runs directly counter to the evidence collected by the Bureau of Alcohol, Tobacco and Firearms' Project Identification.
Seventy-one percent, or 7,538, of the handguns submitted for tracing, had a barrel length of 3 inches or less. Sixty-one percent, or 6,476, had a caliber of .32 or less. Since both of these factors relate to the size of the weapon, these figures indicate that concealability is an overriding factor in selecting a handgun for use in crime. [97]
Sawed-off shotguns will be much longer and much bulkier than any of these short and small-caliber handguns, especially "Saturday Night Specials," which combine a caliber of .32 or less with a barrel length of three inches or less, comprised 44% of all the weapons successfully traced, and fit into the palm of an average sized hand. We may conclude, then, that because of the difficulty of concealment, neither long guns nor sawed-off versions of the same are likely to be used in great numbers to replace handguns in the commission of crimes. [Page 272]
The difficulty of concealment factor will outweigh the greater lethalness of long gun shots. Consequently, a ban on handguns will indeed result in a decrease in firearms-related homicide and other violent crimes. Since firearms are the most lethal weapons, and they were used in 64.1% of homicides in the United States in 1990, [98] such a ban is, therefore, likely to result in a reduction in the overall murder rate. [99]
E. Defensive Uses of Handguns
1. General Crime Reduction
The anti-ban literature which I have so far discussed tries to show that permissive handgun laws do not increase homicide and violent crime. I have argued that these arguments fail. Opponents of a handgun ban are at their strongest, however, when they discuss alleged positive benefits of handgun ownership. They focus, in particular, on the value of guns in self-defense, and in defending property against robbery and theft. Any benefits of my proposal in reducing crime would have to be weighed against the reduction in defensive uses of handguns that would presumably result from a handgun ban. [100] These arguments are based on common sense observations about the likely effectiveness of handguns (as opposed to other weapons, or no weapons at all) in self-defense, and the deterrent effect that this is likely to have on potential assaulters, robbers, and burglars. Situations in which guns would be useful would be assaults and robbery attempts on the street, as well as assault and burglary attempts in the home. This deterrence is over and above any deterrence that may be exerted by the fear of apprehension and arrest by the police.
An often-quoted study intended to establish this deterrent [Page 273] effect was done by Wright and Rossi. [101] They interviewed over 1800 prisoners in ten States concerning their attitude toward the possibility of armed victims. 34% said that they had been "scared off, shot at, wounded, or captured by an armed victim"; another 34% said that they were concerned that they might be shot by their victim); (interestingly, the same percentage feared being shot by the police); 57% said that "most criminals are more worried about meeting an armed victim than they are about running into the police"; and 59% agreed that "a store owner who is known to keep a gun on the premises is not going to get robbed very often." [102] While no one would deny that firearms do have a deterrent effect on potential felons, the reliability of a survey conducted among felons is very dubious. Referring to a similar survey done by Wright and Rossi, a supporter of the defensive use of firearms pointed out the difficulties in relying on surveys of convicted criminals (who, as a group, are remarkable neither for honesty nor acute introspection). Then there are the difficulties in extrapolating from their answers to the attitudes of fellow criminals who, perhaps because of distinguishing characteristics such as greater shrewdness, have not been caught. [103]
Another widely-quoted study was performed by Gary Kleck. [104] Based on six surveys, he estimates that handguns are used 645,000 times per year in defensive response to crimes or attempted crimes. This allegedly compares with 580,000 "criminal misuses" of handguns per year. [105] Moreover, he estimates that "gun wielding civilians in self-defense or some other legally justified cause" kill between 1,527 and 2,819 felons in 1980. [106] To be fair, Kleck does stress that the most common defensive use of handguns is achieved without firing a shot. Merely waving a gun at the aggressor or intruder is often enough to scare him off. Moreover, he envisages a generalized deterrent effect of handguns, whereby the well-known, widespread ownership of guns makes a criminal hesitate before committing any assault, robbery, or burglary, for fear of retaliation.
Kleck's estimate of the number of self-defensive uses of guns is a projection based on surveys, and is subject to a serious criticism. [Page 274]
The respondents were gun owners who have a vested interest in exaggerating both the need for self-defense, and the effectiveness of their guns in providing it. What gun owner, for instance, is going to admit that he fired his gun at a false alarm, or that he used more force than necessary in repelling an intruder who turned out to be harmless? Such nuances are likely to be left out of responses to surveys, and the incidents in question will be recorded as successful defenses of self and property. Even more likely than deliberate dishonesty among respondents to surveys is self-deception and outright error concerning the need to use a gun in self-defense. A chilling example of this trigger-happy attitude is provided by an incident involving Bernard Goetz before he achieved notoriety as the "subway vigilante." At 8pm. one evening he was asked for money by "a crazy kid on drugs" who was walking behind him on Sixth Avenue. Even though he admits that there were many other ways to deal with the situation, Goetz pulled his gun on the youth. [107] In apparent support of his advocacy of the defensive efficacy of handguns, Kleck cites the infamous subway shootings of 1984, by referring to the sharp decrease in subway crime which followed them. [108] Not only must unjustifiable uses of firearms, when less force would have been sufficient to escape the real or perceived danger, be subtracted from the benefits that Kleck claims for the defensive use of guns; they must be added to the long list of bad consequences of handgun ownership around which this paper is based.
Kleck's estimate of 1,527 to 2,819 self-defensive killings in 1980 is based on his claim that the official number of justifiable gun homicides by civilians reported to the FBI (379) vastly underestimates the real number. His projection is based on an extrapolation from data from Detroit and Miami. One assumes that the vast majority of these justifiable homicides were responses to the threat of murder. Let us generously assume that all 2,819 of Kleck's self- defensive killings did indeed prevent killings. In the same year, there were at least 13,650 murders with firearms. [109] Kleck has failed to prove that the self-defensive use of firearms outweighs their abuse in homicide.
The most impressive evidence quoted by Kleck consists of statistics indicating that resisting robbery and assault with firearms is the best way to minimize the chances of being injured or robbed. Only 30.9% of robberies are completed against armed victims, of [Page 275] whom only 17.4% are injured; and only 12.1% of armed victims are injured in assaults. Interestingly, the next best strategy is non-resistance. While 88.5% of robberies are completed against them, only 24.7% of nonresistors are injured; and only 27.3% of nonresistors are injured in assaults. [110] Armed victims, then, appear to be less likely to be robbed or injured than nonresistors. However, it must be remembered that Kleck's source is another survey, this time of victims of attempted assaults and robberies, and that, as pointed out above, gun owners are likely to exaggerate the defensive value of their weapons.
More generally, a substantial number of the crimes of the kind that Kleck alleges are prevented by the defensive use of firearms are themselves committed with guns (64.1% of homicides, 36.6% of robberies, and 23.1% of aggravated assaults.) [111] Even if Kleck is right that a ban on handguns would reduce people's ability to defend themselves, it would also reduce the need for self-defense in the first place. A heavily-armed citizenry might be a rational response if heavily-armed criminals were inevitable; but far more rational would be a society that strives to disarm all private citizens, thus obviating the need to use firearms in self-defense.
The reasoning that seeks safety in the proliferation of privately owned firearms is precisely the rationale that supported nuclear proliferation under the strategy of mutual assured destruction (MAD). This policy rested the survival of the human race on the hope that mutual fear of retaliation would prevent a first strike. It has been heavily criticized on the ground that an unspeakable catastrophe could follow an accidental firing of a nuclear missile, or a deliberate attack by a fanatical nation that did not care about retaliation. These criticisms parallel those that I have levelled at the argument for handguns as self-defense, with the difference that lethal accidents with and aggressive abuse of handguns are an everyday reality, rather than a feared possibility. A further parallel is that the high rate of handgun ownership in this country is self-perpetuating. First, it is in response to the proliferation of handguns that an increasing number of people believe they need to buy a handgun for self-defense (though, as I have argued, it is an illusion that more widespread ownership of guns will decrease gun crime.) Second, while some potential criminals may be deterred by a heavily-armed citizenry, others will arm themselves with more and more powerful firearms in order to outgun resisters. Trading gunfire or playing chicken with increasingly heavily-armed criminals is a tenuous basis [Page 276] for the defense of society.
A study by Arthur Kellermann and Donald Reay undercuts the central factual claim made by Kleck in his study. [112] Unlike Kleck's projections, this study is based on an analysis of the actual homicide figures for King's County, Washington (population 1,270,000) from 1978 to 1983. Kellermann and Reay focused on the 398 shooting deaths that occurred in the residence where the lethal firearm was kept. Thus their study gives a useful measure of the value of handguns (which comprised over 70% of the firearms involved) as defensive weapons when kept in the home. The authors conclude somewhat misleadingly that there were "53 suicides, criminal homicides, or accidental gunshot deaths involving a gun kept in the home for every case of homicide for self-protection." [113] They achieve this startling 53:1 by confining their attention to only "self-protection homicide." When other kinds of defensive homicide are also considered, the ratio between suicides, criminal homicides, and accidental deaths and, on the other hand, defensive killings, becomes 21:1.
As Kellermann and Reay themselves note, the defensive efficacy of guns is not exhausted by the number of people killed in this way. As noted above, people with guns are most successful when they frighten off the would-be criminal before he even attempts a crime; and the study measures only defensive homicides, not defensive uses of firearms which prevented crimes by causing nonlethal injuries or by causing the criminal to flee. They also recognize the limitations of a study which focuses on just one county, which might turn out to be a freakish, unrepresentative case.
Another qualification which is in order is to point out the key role played by suicides in producing the outlandish ratio quoted in the study. [114] Suicide rates throughout the world are far higher than homicide rates, and it would be unfair to lay the blame for these deaths at the doorstep of handguns. The suicide rate in many countries with stricter handgun control than the United States is higher than that in the United States. [115] Nonetheless, Kellermann and Reay are justified in asserting "given the high case-fatality rate associated with suicide attempts involving firearms, it seems likely that easy access to guns increases the probability that an impulsive [Page 277] suicide attempt will end in death." [116] A more pertinent comparison which can be drawn from their evidence about firearms deaths is that alongside the eighteen defensive homicides that occurred in King County during the six year period, there were fifty-three criminal homicides and accidental deaths. Instead of Kellermann and Reay's ratio of 53:1, we have one of 2.94 criminal homicides and accidental deaths for every defensive homicide resulting from firearms. It must be remembered, moreover, that any number of these defensive homicides may themselves have been precipitated by the offensive use of firearms.
The value of their study is to provide concrete evidence, based on actual gunshot wounds, of the startling disparity between the proven lethal abuses of firearms and the proven defensive uses. It does not preclude the kind of defensive efficacy of handguns asserted by Kleck, and indeed it is hard to deny that the threat of deadly retaliation will sometimes be a powerful deterrent to crime. However, it certainly puts the burden of proof on him to produce more evidence, beyond his projections from statistics and extrapolations based on surveys, that the positive effects of permissive handgun laws outweigh the negative effects demonstrated by the study. Kleck himself claims only that the evidence he adduces is "compatible with" his hypothesis. [117] A major theme of this paper, reinforced by Kellermann's and Reay's study, is that the onus is on defenders of handguns to show that their thesis is in fact true, not that it may be true. In the meantime, it is fair to conclude that Kellermann and Reay provide strong reason to doubt whether the defensive value of handguns kept in the home outweighs their dangers.
It cannot be denied that owning handguns may make people feel secure in their homes and on the streets. However, if my arguments in this section are sound, this feeling is illusory, and an illusory feeling of security cannot seriously be advanced as more important than a substantial reduction in murder and violent crime.
Taking handguns from law-abiding citizens does not deprive them of many methods of self-defense. They still have the option of escaping or calling for help, using weapons other than handguns, using their bare hands, reasoning with the criminal, or simply not resisting (which, as I pointed out above, is the next best way to avoid being injured.) It is possible that in some cases a victim would have been able to avoid theft, injury, or even death had she been armed with a handgun. This "cost" of my proposal needs to be weighed against the likely negative results of the defensive use of [Page 278] handguns described above: unnecessary and excessive use of handguns in self-defense; and the deaths shown by Kellermann and Reay to result from the abuse of handguns in the home.
An even more important set of objections to the defensive use of handguns is beyond the scope of the largely utilitarian argument of this paper. Can allowing private citizens to effectively become both the arbiter of law and the administrator of justice be reconciled with the rule of law in a civilized society? Don Kates traces a long history of private law enforcement in both Britain and the United States, [118] and Bruce Benson argues for its effectiveness. [119] However, neither of them consider the serious moral issues concerning the extent to which force may justifiably be used defensively, and the propriety of allowing private citizens to act as law enforcers.
It should be remembered that these drawbacks pertain to the defensive use of handguns, which is put forward by opponents of a handgun ban as one of the strongest reasons for allowing private citizens to own handguns. Elsewhere in my paper I have detailed the many other reasons for banning handguns. This section has given little reason to believe that any lives saved through the defensive use of handguns would outweigh the reduction in the current annual handgun murder toll of 9,923, and in the number of handgun assaults and robberies, that would be effected by a handgun ban.
2. Handguns and Violence against Women
The case for self-defensive ownership becomes stronger, however, in the context of violent crimes against women. [120] Guns take on special significance in the case of crimes committed by men against women because they enable women to compensate for the greater physical strength of men. [121] A woman armed with a handgun will be able to repel an attack by a much stronger male, unless [Page 279] he too is carrying a firearm. As was claimed in the previous section on self-defense in general, the value of handgun possession by women is not to be measured in terms of the number of assailants whom they kill, wound, or repel by shooting at them. Instead it manifests itself in the deterrent effect produced by the knowledge that a particular woman owns a gun, and the general deterrence created by wide-spread female ownership of handguns. Both kinds of deterrence would minimize the number of women who have to actually use their guns. Silver and Kates approvingly cite the slogan, "God didn't make men and women equal, Colonel Colt did." [122] While one may want to avoid basing women's moral equality on their possession of lethal firepower, anything which offers even prima facie promise of reducing violence against women deserves serious consideration.
Unless it was clear that she could have escaped without killing the assailant, the use of lethal force by a woman in order to protect herself from murder, rape or aggravated assault will rarely be criticized. The apparent inability of the police to protect women from threats of violence is illustrated by graphic cases in which police have insisted that they cannot arrest an assailant unless he has actually initiated aggression, thus strengthening the case for private self-protection. [123] The fact that women are responsible for such a small percentage of violent crime [124] indicates that the protection afforded women by handguns is unlikely to be accompanied by an increase in handgun crime.
A major theme of this paper has been that the benefits of self-defense from handguns are in general outweighed by the danger of handgun abuse. Since the vast majority of handgun abuse is by men, a suggested policy is a ban on male handgun ownership, while allowing women with a proven need for self-protection to own handguns. (In light of the high rate of violence against women, perhaps all women have this proven need.) However, such a sex-based ban would not be feasible. First, men could easily solicit women to "buy for" them. Second, a policy that discriminates against males would be subject to constitutional challenge. It appears, then, that the only way to allow women the defensive benefits of handguns is to make handguns available to both women and men.
The case for female ownership of handguns is highlighted by [Page 280] an examination of spousal homicide. From 1977 until 1984, just over one half of spousal homicides were committed by husbands, [125] and this number had increased to 70.6% by 1991. [126] More important, a significant proportion of wife-husband homicides are committed by abused wives who are defending themselves or their children. [127] More widespread ownership of handguns appears likely to reduce the number of wives killed by their husbands. While it would be preferable for women to protect themselves without killing their husbands, there can be little objection to their doing so if this is the only way for them to save their own lives. To eliminate handguns altogether may reduce the number of wife-husband killings, but it "would only change the sex of the decedents by ensuring that, in virtually every case, it would be the abused wife, not the murderous husband" who died. [128] Other things being equal, it seems fair that the aggressive husband, and not the victimized wife, be the one to forfeit his life. While it is true that a woman's handgun will be of less defensive value against an armed assailant, the fact is that women's assailants rarely use firearms.
In 89.6% of the violent crimes directed against women during the 10 years of 1973-82, the offender did not have a gun; only ten percent of rapists used guns and only twenty-five percent of nonstrangers who attacked victims (whether male or female) had any weapon whatever. [129]
After an extensive study of the "spouse abuse literature", Howard concedes that reducing ownership of firearms may actually decrease spousal homicide:
To the extent it is a crime of passion, spousal homicide should respond to enforced legislation banning all guns ... If a [handgun only] ban removed handguns from the reach of angry husbands and wives, many of these individuals would simply grasp the most conveniently available alternative weapon, as the data on spousal homicide suggests they are already doing. The substituted weapon would probably [Page 281] be a knife and increased use of knives might decrease the number of fatalities, given the lower lethality of knives as compared with guns. [130]
However, Howard recommends "increasing the options available to battered wives" (e.g. tougher laws against spouse abuse, more shelters for victims of domestic violence) as a better method of protecting women than gun control. She advocates attacking the psychological roots of violence against women, instead of focussing on handguns, which are simply one of the exacerbating factors. Because of the argument that physically weaker women need handguns to protect themselves against whatever violence would still occur, she opposes a handgun ban.
The case for allowing women to own handguns, then, hinges on their role as an "equalizer" to compensate for men's superior strength. I do not dispute that they are sometimes sufficient for doing so. My point is that they are not necessary, since there are alternative ways to protect women. [131] With regard to current or former husbands or boyfriends who threaten violence, restraining orders and police protection can make a difference, though the latter will require substantial public funding. Changing locks, and installing secure doors and burglar alarms can make homes more secure. Martial arts and other forms of unarmed self-defense can be highly effective against an assailant without a firearm, yet these options are dismissed summarily by Silver and Kates, on the grounds that guns are (1) "less arduous," and (2) more effective. [132]
Few would dispute that guns are more effective; but if unarmed self-defense is sufficiently effective in warding off attacks, using a more lethal method of defense would be gratuitous violence. Nor need we confine our attention to unarmed resistance. Mace is a very effective weapon which can be used to immobilize an assailant, without causing serious injury. It also has the advantages of being even easier to conceal and use than a handgun, and of being far less likely than handguns to be used in the commission of crimes. Being neither lethal nor capable of causing permanent injuries, it will be less effective in intimidating victims into submission. Against the 10% or so of women's assailants who do carry firearms, mace would be less effective, but it has to be realized that handguns are also of less use against an assailant with firearms. Moreover, my proposed [Page 282] handgun ban would make it less likely that he would be armed in the first place. In contrast, a likely result of women carrying handgun is the same proliferation described in the previous section: more of their assailants would carry guns, in order to ensure the success of their attack.
Let us concede for the sake of argument that the combination of all of the alternative methods of self-defense that I have proposed would still be marginally less effective than handguns in protecting women against violence in the 10% or so of assaults that involve firearms. The inference to the conclusion that a handgun ban would decrease protection for women results from a comparison between a world in which handguns are banned, and an imaginary world in which most women arm themselves with handguns. In the actual world women may now legally own handguns, but the vast majority choose not to do so.
The relevant comparison is between the actual world, in which handguns are used in hundreds of thousands of violent crimes every year, yet in which few women own handguns; and, on the other hand, a world in which a handgun ban substantially reduced the number of handguns owned by both women and their potential assaulters. Whatever protection would be lost by disarming the small number of women who currently own handguns is outweighed by the reduction in violence against women that would be effected by a handgun ban, which would take one of the most potent weapons out of the hands of many potential assaulters.
It is true that 50% of those who own guns solely for defense are female. [133] However, far more men than women own guns. [134] Given women's extra vulnerability, and the fact that there are now many more female-headed households than in the mid-sixties, one would expect more women to own guns. In fact, a Harris poll showed that gun ownership in female-headed households was less than a half of that in homes in which an adult male lived. [135] The indications are that women themselves, whatever their extra vulnerability may be, are generally unconvinced of the need to own handguns for self-defense.
The alleged protection for women resulting from the defensive ownership of handguns, then, fails to provide a serious objection to a handgun ban. In contrast, throughout my paper I have detailed the substantial reduction in murder and violent crime that is likely to [Page 283] result from a handgun ban. Women, too, are the beneficiaries in a society in which far fewer of their loved ones are killed and maimed.
V. CONCLUSION
After a detailed analysis of the literature that opposes a handgun ban, I have shown that none of these arguments seriously respond to the "burden of proof" challenge which I present in section IC. I have established a strong prima facie case for my hypothesis, justifying at least an experimental handgun ban, for, say, twenty-five years. [136]
If my hypothesis is wrong, a minor restriction on people's behavior will have been needlessly imposed, and whatever self-defense handguns may have provided will have been lost. This loss is minimal in comparison with the many harmful uses of handguns which, if I am correct, would be prevented by a handgun ban. Consequently, even assuming that there is only a 50% chance that my hypothesis is true (though I have argued that the probability is far higher), a handgun ban is justified on the ground of its greater expected utility.
I have not addressed what may be considered the strongest objection to a handgun ban: the Second Amendment, and its guarantee of the right to bear arms. What I have shown is that there is a strong utilitarian case for banning handguns, and that the constitutionality of such a ban therefore merits careful consideration. [137]
* Associate Professor of Philosophy, Alma College, Alma, Michigan.
[1]. See, e.g., Don B. Kates, Jr., Handgun Prohibition and the Original Meaning of the Second Amendment, 82 MICH. L. REV. 204-73 (1983); Sanford Levinson, The Embarrassing Second Amendment, 99 YALE L.J. 637-59 (1989-90); Wendy Brown, Guns, Cowboys, Philadelphia Mayors, and Civic Republicanism: On Sanford Levinson's 'The Embarrassing Second Amendment', id. at 661-67.
[2]. FEDERAL BUREAU OF INVESTIGATION, UNIFORM CRIME REPORTS 12 (1990). The actual number of handgun murders is higher, since 9,923 is the number that emerged from an analysis of just 85.5% of the total murders.
[3]. Id. at 21.
[4]. Id. at 24.
[5]. Id. at 12; FRANKLIN E. ZIMRING & GORDON HAWKINS, THE CITIZEN'S GUIDE TO GUN CONTROL 38 (1987).
[6]. The danger that long guns will be substituted for handguns in the event of a handgun ban will be discussed infra. section II.D.
[7]. In Canada, handgun owners are allowed to keep their weapons at home, and then transport them to shooting ranges only in a locked box. Criminal Code of Canada, Firearms and Other Offensive Weapons, MARTIN'S CRIMINAL CODE OF CANADA, § 81-016.9 (1982). Because of the far greater rate of handgun violence in the United States (see infra text at 248 for comparative handgun homicide rates), target shooters must not be allowed to keep handguns at home in this country.
[8]. James J. Jacobs, Exceptions to a General Prohibition on Handgun Possession: Do They Swallow Up the Rule?, in LAW AND CONTEMPORARY PROBLEMS 49, (Don B. Kates, Jr. ed., 1986).
[9]. Id. at 7-13.
[10]. Id. at 7.
[11]. Id. at 18.
[12]. Id. at 6.
[13]. Jacobs, supra note 8, at 33.
[14]. Id. at 14-22. His arguments are, incidentally, very damaging to those who oppose a handgun ban. If even the police can do their job without the use of guns, then the claim of the average citizen to need a handgun for self-defense is greatly weakened.
[15]. A tragic instance involved a store owner who shot and killed a 13 year-old intruder whom he had already immobilized by hitting him on the head. See Owner Charged in Slaying at Store, DETROIT FREE PRESS, Sept. 2, 1991, at A1.
[16]. THE NEW ENCYCLOPAEDIA BRITANNICA (1990); except for Australia's figure, which was supplied by the Australian Institute of Criminology. Correspondence from Dr. J.M. Herlihy, Australian Institute of Criminology (Nov. 25, 1991).
[17]. A press release from the Bureau of Alcohol, Tobacco and Firearms, May 22, 1991, estimates the number of handguns either privately owned or else "available for sale" at 66,666,000. Gary Kleck, a supporter of private handgun ownership, estimates that by 1978 there were 47 million privately owned handguns. Gary Kleck, Guns and Self-Defense: Crime Control Through the Use of Force in the Private Sector, in SOCIAL PROBLEMS 35:1, 4 (Feb. 1988). Since the data for other countries that I use concerns not handguns available for sale, but those actually privately owned, taking the average of these numbers seems a reasonable compromise, which if anything underestimates the number of handguns actually owned by now.
[18]. The Office of Information Services and International Relations of the Israeli Police Department reports that in 1990 there were 171,448 licensed handguns owned by civilians in Israel. Correspondence from Naomi Shapira, Information Services of International Relations (Dec. 15, 1991). Since all of the other estimates in my table refer to the total number of handguns, legal and illegal, (a figure that the Israeli Police Department was unable to supply), the number which I have quoted for Israel is almost certainly an under estimate.
[19]. The Swedish Central Police Agency (Rikspolisstyrelsen) estimates that there are 2.5 million firearms owned by civilians in Switzerland. Correspondence from Staffan Thunqvist, National Swedish Police Board (Nov. 25, 1992). I arrived at my estimate of the number of handguns by dividing this number by 8.11, which is the average ratio of total firearms to handguns in the countries for which I have this information (Australia, Canada, Great Britain, and the United States).
[20]. A Canadian Justice Department study estimates that there are 5.9 million privately owned firearms in Canada. See generally ANGUS REID GROUP, INC., FIREARMS OWNERSHIP IN CANADA 4 (1991). Of this number, the Justice Department estimated that 440,000 were handguns. James Hayes, the Coordinator of the Canadian Department of Justice Firearms Control Task Group, now estimates the number of handguns at 750,000. Telephone Interview with James Hayes, Canadian Justice Department (Feb. 22, 1993). I have used the mid-point of these two estimates of the number of handguns in Canada.
[21]. In 1987 the Australian Institute of Criminology estimated that there were at least 3.5 million guns of all kinds in Australia, legal and illegal. To estimate what number of these guns are handguns, I refer to the results of an International Crime Victims Survey (1990), which included Australia. It found that 20.7% of Australian households own a gun of some kind, while 1.6% own a handgun. Thus 7.72% of gun-owning households own handguns. See Duncan Chappell, A National Gun Control Strategy: The Recommendations of the National Committee on Violence, Address Before the Australian Crime Prevention Council National Conference, Wollongong (Sept. 4, 1991).
An independent measure of the total firearms to handguns ratio can be obtained by comparing the number of licensed handguns and long guns in the four states/territories for which these numbers are available from the Australian Institute of Criminology (Victoria, South Australia, Tasmania, and Northern Territory). Handguns on average comprise 7.36% of the licensed firearms in these jurisdictions. See Anita Scandia, "Numbers of Licensed Firearms and Shooters by State and Territory," Australian Institute of Criminology document (Aug. 29, 1991). My estimate of the total number of handguns privately owned in Australia is obtained by applying the mean of these two very similar, but independently derived measures (7.54% of all firearms are handguns) to the estimated total of 3.5 million firearms.
[22]. Michael Black reports that there are 160,000 licensed holders of "firearms" (handguns) in Britain. Gun Law in Britain: How Hard Is It to Buy a Gun? ILLUSTRATED LONDON NEWS, Aug. 1988 at 24. He cites estimates that the actual number of handguns "in legitimate circulation" is three times higher, (i.e. 480,000.)
[23]. For details on handgun and other firearms laws in other countries, see LAW LIBRARY OF CONGRESS, FIREARMS REGULATIONS IN VARIOUS FOREIGN COUNTRIES (1990).
[24]. The historical interest in gun ownership in the United States is evidenced by the Second Amendment's guarantee of the right to bear arms. Further evidence of the high demand for handguns is provided by current ownership rates.
[25]. This elementary point is persistently ignored by the Tobacco Institute. The institute still uses thought experiments to deny that smoking tobacco causes lung cancer despite the overwhelming evidence in favor of this hypothesis. A causal hypothesis is not refuted by showing that it is not logically true, which is a condition that no informative hypothesis can meet.
[26]. See Ernest van der Haag, On Deterrence and the Death Penalty, 60 JOURNAL OF CRIMINAL LAW, CRIMINOLOGY, AND POLITICAL SCIENCE (1969), reprinted in RAZIEL ABELSON & MARIE-LOUISE FRIQUEGNON, ETHICS FOR MODERN LIFE 208 (1982).
[27]. For a discussion of this argument, see David Conway, Capital Punishment and Deterrence: Some Considerations in Dialogue Form, 3 PHILOSOPHY AND PUBLIC AFFAIRS (1974), reprinted in JOHN ARTHUR AND WILLIAM H. SHAW, READINGS IN PHILOSOPHY OF LAW 258-59 (1984).
[28]. This strategy is the same as that used by capital punishment advocates in order to explain the recalcitrant evidence. See supra text section I.D.
[29]. FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY 529 (Don B. Kates, Jr. ed., 1984).
[30]. An exception is Gary Kleck, who devotes one and one-fourth pages to advocating renewed emphasis on social programs such as the Job Corps. Kleck claims such measures help address some of the underlying causes of crime. See Gary Kleck, Policy Lessons From Recent Gun Control Research, in LAW AND CONTEMPORARY PROBLEMS 49, 61-62 (Don B. Kates, Jr. ed., 1986).
[31]. See RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 31-32 (Don B. Kates, Jr. ed., 1979); Bruce L. Benson, Guns for Protection and Other Private Sector Responses to the Fear of Rising Crime, in FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY 351-52 (Don B. Kates, Jr. ed., 1984).
[32]. See ZIMRING AND HAWKINS, supra note 5 at 8-9.
[33]. See RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 38 (Don B. Kates, Jr. ed., 1979).
[34]. 26 ENCYCLOPEDIA AMERICANA 148 (1991).
[35]. John Henry Sloan, Arthur L. Kellermann, Donald T. Reay, James A. Ferris, Thomas Koepsell, Frederick P. Rivara, Charles Rice, Laurel Gray, and James LoGerfo, Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities 319 NEW ENG. J. MED. 1256-62 (1988).
[36]. These two cities are particularly well chosen to eliminate any bias due to the generally higher level of violence, especially homicide, in the United States as compared with Canada. Seattle's homicide rate is only 50-70% that of other major United States cities, while Vancouver's homicide rate is two to three times higher than that of Ottawa, Toronto, and Calgary. Id. at 1259.
[37]. Id. at 1256-7.
[38]. Id. at 1259.
[39]. Id.
[40]. Id.
[41]. JOHN HENRY SLOAN, ET AL., Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities 319 NEW ENG. J. MED. 1256, 1259 (1988).
[42]. Id. at 1259.
[43]. Id.
[44]. If one excludes those homicides committed by the police in the line of duty, and all those homicides done in self-defense or otherwise deemed justifiable (32 in total), the relative risk of being murdered in Seattle is still 1.57 that in Vancouver. Id.
[45]. Id. at 1257.
[46]. JOHN HENRY SLOAN ET AL., Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities 319 NEW ENG. J. MED. 1256, 1257 (1988).
[47]. Id.
[48]. Id.
[49]. Id. at 1258.
[50]. Id.
[51]. JOHN HENRY SLOAN ET AL., Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities 319 NEW ENG. J. MED. 1256, 1258 (1988).
[52]. James Wright, Guns and Sputter, REASON, 46-47 (July 1989). See also Paul H. Blackman, Medical Journal's Article Seriously Flawed, NRA Says, NEW ENGLAND JOURNAL OF MEDICINE, reprinted in NRA OFFICIAL JOURNAL, 55-56 (Jan. 1989). See also 320 THE NEW ENGLAND JOURNAL OF MEDICINE, 1214-17 (May 4, 1989). In order to keep discussion managable, I will focus on the two objections I have listed, which I consider to be the most telling.
[53]. James Wright, Guns and Sputter, REASON, 46-7 (July 1989).
[54]. Id.
[55]. Id.
[56]. SLOAN ET AL., supra note 35 at 1260.
[57]. Id.
[58]. Id.
[59]. Id.
[60]. Support for this hypothesis is provided by a study that is favorable to the view that race is a major determinant of the homicide rates in American and Canadian cities. See Robert J. Mundt, A Tale of Four Cities: Firearms and Violence in the U.S. and Canada, Remarks at the Annual Meeting of the American Political Science Association, (Aug. 29, 1991). Mundt concludes (23-6) that race alone does not account for the higher rate of crime in the American cities he studies and that the greater availability of firearms may also be a causal factor.
[61]. In 1991, 51% of murder victims and 55% of known murderers in the United States were black. UNIFORM CRIME REPORTS 16 (1991). Only 12.3% of the population is black. STATISTICAL ABSTRACT 16 (1992).
[62]. MUNDT, supra note 60 at 25.
[63]. See, e.g., Mark K. Benenson and David T. Hardy, Critiquing the Case for Handgun Prohibition, in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 81-82 (Don B. Kates, Jr. ed., 1979). See also Don B. Kates, Jr. and Mark K. Benenson, who conclude that "tested over a wide spectrum of demographically, culturally, and geographically diverse states, handgun prohibition laws nowhere appear to have had any reductive effect upon any type of violent crime." Handgun Prohibition and Homicide: A Plausible Theory Meets the Intractible Facts, in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 94 (Don B. Kates, Jr. ed., 1979).
[64]. Id. at 96-100.
[65]. THE DEPARTMENT OF TREASURY, PROJECT IDENTIFICATION: A STUDY OF HANDGUNS USED IN CRIME (May 1976).
[66]. Id. at 19.
[67]. Id. at 14.
[68]. Id. at 14-15.
[69]. Id. at 14.
[70]. THE DEPARTMENT OF TREASURY, PROJECT IDENTIFICATION: A STUDY OF HANDGUNS USED IN CRIME (May 1976).
[71]. The results of the study done by the Bureau of Alcohol, Tobacco and Firearms also serve as a response to the objection that gun laws are unenforceable.
[72]. A similar response applies to another utilitarian objection that refers to the likelihood of illegal and "homemade" gun production in this country in the event of a handgun ban. Restrictions on when guns may be owned legally are quite compatible with a crackdown on illegally made guns. Moreover, the fact that such guns are inaccurate and dangerous to the user will also act as a restraint on illegal gun production.
[73]. See David Hardy and Don Kates, Handgun Availability and the Social Harm of Robbery: Recent Data and Some Projections, in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 129-30 (Don B. Kates, Jr. ed., 1979).
[74]. See KLECK, supra note 30 at 41.
[75]. KATES, supra note 29 at 528.
[76]. However, it should be noted that the very arguments raised against a general prohibition on handguns (i.e. that criminals are unlikely to respond to it) also make it unlikely that a restriction targeted at felons will fare any better.
[77]. KLECK, supra note 30 at 41-43.
[78]. UNIFORM CRIME REPORTS 13 (1990).
[79]. Id.
[80]. Id. at 14.
[81]. Gary Kleck and David J. Bordua, The Assumptions of Gun Control, in FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY 34-43 (Don B. Kates, Jr. ed., 1984).
[82]. James D. Wright, Second Thoughts About Gun Control, 91 THE PUBLIC INTEREST 31 (Spring 1988).
[83]. KLECK AND BORDUA, supra note 81 at 42.
[84]. Id. at 43.
[85]. 7442 per annum, based on the FBI's handgun homicide numbers. UNIFORM CRIME REPORTS 12 (1990).
[86]. Sentence enhancement adds a mandatory prison term onto the sentence of any felon convicted of a crime that involves the use of a firearm. See Alan Lizotte and Marjorie S. Zatz, The Use and Abuse of Sentence Enhancement for Firearms Offenses in California, LAW AND CONTEMPORARY PROBLEMS 49, 199-221 (Don. B. Kates, Jr. ed., 1986).
[87]. This bill, which resulted from amendments to the proposed Brady Bill, requires a five day waiting period during which a background check is conducted. The applicant may then purchase a handgun if and only if this check shows that he has no criminal record. The waiting period will be phased out when a national computerized instant background check system becomes available.
[88]. Zimring and Hawkins have shown that approximately five times as many gun assaults as knife assaults result in death. See ZIMRING AND HAWKINS, supra note 5, at 15. In response, Hardy and Kates argue that criminals will likely substitute ice picks and butcher knives, which are almost as lethal as handguns. HANDY AND KATES, supra note 73 at 123-25. The mind boggles at how such weapons are supposed to be concealed before robberies. Also in response, Gary Kleck has argued that the death rates found by Newton and Zimring may be due not to the intrinsic lethalness of the guns themselves, but "the greater seriousness of intent to injure or kill among users of guns." KLECK, supra., note 30 at 38. Both the "ice pick and butcher knife" and the "serious intent" conjectures are compatible with Zimring's and Hawking’s evidence. However, no evidence is given to show that either of these intuitively implausible hypotheses is actually true. See generally ZIMRING AND HAWKINS, supra note 5 at 13-21.
[89]. HARDY AND KATES, supra note 73 at 121-22.
[90]. Id. at 122-23.
[91]. Don B. Kates and Mark K. Beneson, Handgun Prohibition and Homicide: A Plausible Theory Meets the Intractible Facts, in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 111 (Don B. Kates, Jr., ed., 1979).
[92]. Id.
[93]. KLECK, supra note 30 at 49.
[94]. 9.88%, according to the UNIFORM CRIME REPORTS 12 (1990).
[95]. KATES AND BENESON, supra note 91 at 114.
[96]. HARDY AND KATES, supra note 73 at 127.
[97]. PROJECT IDENTIFICATION, supra note 65 at 2.
[98]. UNIFORM CRIME REPORTS at 12.
[99]. Despite my scepticism about the "long gun substitution theory" which is given in objection to handgun bans, I fully support one proposal that is made in the light of this theory: we need to restrict long guns in order to minimize criminal substitution of such guns in the event of handgun restrictions. See, Gary Kleck, Handgun-Only Gun Control: A Policy Disaster in the Making, in FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY 197-99 (Don B. Kates, Jr., ed., 1984). A background check on purchasers of long guns (of the kind recently proposed for handgun purchasers) could be profitably combined with a ban on handguns.
[100]. James D. Wright, The Ownership of Firearms for Reasons of Self- Defense, in FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY 301-27 (DON B. KATES, JR. ED., 1984); KLECK, SUPRA NOTE 30 AT 43-48; DON B. KATES, JR., GUNS, MURDER, AND THE CONSTITUTION 17-36 (Pacific Research Institute for Public Policy) (1990).
[101]. KLECK, supra note 30 at 46.
[102]. Id.
[103]. Daniel D. Polsby, Reflections on Violence, Guns, and the Defensive Use of Lethal Force in LAW AND CONTEMPORARY PROBLEMS 49, 97 (Don B. Kates, Jr. ed., 1986).
[104]. Gary Kleck, Guns and Self-Defense: Crime Control Through the Use of Force in the Private Sector, 35 SOCIAL PROBLEMS 4 (Feb. 1988).
[105]. Id. at 4.
[106]. Id. at 5.
[107]. George P. Fletcher, A CRIME OF SELF DEFENSE: BERNHARD GOETZ AND THE LAW ON TRIAL, 104-5 (1988).
[108]. KLECK, supra note 104 at 15.
[109]. UNIFORM CRIME REPORTS 12 (1980).
[110]. KLECK, supra note 104 at 7-9.
[111]. UNIFORM CRIME REPORTS 12,21,24 (1990).
[112]. Arthur Kellermann and Donald Reay, Protection or Peril? An Analysis of Firearm-Related Deaths in the Home, 314 NEW ENGLAND JOURNAL OF MEDICINE, 1557-60 (1986).
[113]. Id. at 1560.
[114]. 37 of the alleged 43 deaths for every self-defensive killing are suicides.
[115]. See Don B. Kates, Jr., GUNS, MURDER, AND THE CONSTITUTION 42 (Pacific Research Institute for Public Policy) (1990).
[116]. KELLERMAN AND REAY, supra note 112 at 1559.
[117]. Supra note 104 at 17.
[118]. Kates, supra note 115, at 19.
[119]. Bruce L. Benson, Guns for Protection and Other Private Sector Responses to the Fear of Rising Crime in FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLICY (Don B. Kates, Jr. ed., 1984)
[120]. See Carol Ruth Silver and Don B. Kates, Self-Defense, Handgun Ownership, and the Independence of Women in a Violent, Sexist Society in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 139-69 (Don B. Kates, Jr. ed., 1979); KATES, supra note 115 at 24-32; Margaret Howard, Husband-Wife Homicide: An Essay From a Family Law Perspective in LAW AND CONTEMPORARY PROBLEMS 49, 63-88 (Don B. Kates, Jr. ed., 1986).
[121]. For instance, Kates quotes the finding that "men who batter [their mates] average 45 pounds heavier and 4 to 5 inches taller than the women they attack." KATES, supra note 115 at 24.
[122]. See Carol Ruth Silver and Don B. Kates, Self-Defense, Handgun Ownership, and the Independence of Women in a Violent, Sexist Society in RESTRICTING HANDGUNS: THE LIBERAL SKEPTICS SPEAK OUT 169 (Don B. Kates, Jr. ed., 1979).
[123]. Id. at 144-47.
[124]. In 1991, women committed 10% of homicides, 9% of robberies, and 14% of aggravated assaults, UNIFORM CRIME REPORTS, 16, 29, 34 (1991).
[125]. Margaret Howard, Husband and Wife Homicide: An Essay From a Family Law Perspective in LAW AND CONTEMPORARY PROBLEMS 49, 67 (Don B. Kates, Jr. ed., 1986).
[126]. UNIFORM CRIME REPORTS 19 (1991).
[127]. Kates claims this to be the case in "the overwhelming majority" of wife-husband homicides, without quoting specific evidence giving the percentage. KATES, supra note 115 at 25. Howard is more specific, quoting a survey of husband killers in the Cook County, Illinois jail. 40% of these women had been abused by their victims. She also refers to an earlier study by Wolfgang, in which nearly 60% of the murdered husbands had abused their wives. See HOWARD, supra note 125 at 74-5.
[128]. KATES, supra note 115 at 26.
[129]. Id. at 29.
[130]. HOWARD, supra note 125 at 88.
[131]. As indicated above, Howard suggests several excellent measures to reduce violence against women without using firearms. HOWARD, supra note 125 at 86-87.
[132]. SILVER AND KATES, supra note 122 at 161.
[133]. KATES, supra note 115 at 11-12.
[134]. In the 1960s, only 7% of the people who bought handguns were women. ZIMRING AND HAWKINS, supra note 5 at 186.
[135]. Id. at 187.
[136]. Such a lengthy trial period is necessary in order for a gradual decrease in the vast number of handguns already in circulation to take effect. To this end, I support the "buyback" schemes currently operated by some police departments.
[137]. An early draft of this paper was researched and written at a National Endowment for the Humanities Summer Seminar at the University of Maryland in 1991. I am grateful to seminar members and director David Luban for helpful suggestions, and especially to Sterling Harwood for extensive written comments and discussion.