University of Colorado Law Review
Symposium, Crime is Not the Problem: Lethal Violence in America
Issues Arising from the New Book by Franklin E. Zimring & Gordon Hawkins
vol. 69, no. 4, 1998: 969.
Posted for Educational use only. The printed edition remains canonical. For citational use please visit the local law library or obtain a back issue.

AMERICAN HOMICIDE EXCEPTIONALISM
Daniel D. Polsby *
Don B. Kates, Jr. **

Copyright © 1998 University of Colorado Law Review, Inc., Daniel D. Polsby & Don B. Kates, Jr.

Introduction

The standard account of America's homicide problem, championed in one way or another by many prominent criminologists, has now received its classic statement by Franklin Zimring and Gordon Hawkins. G7 countries, except for Japan, have quite similar crime rates; indeed, their experience with many types of crime seems to have been quite parallel since at least the middle of the twentieth century. [1] All experienced a roughly coincidental surge of larcenies, burglaries, and assaults as each of their post-war baby booms matured into late adolescence; all of them had to cope, and, like America, are still learning to cope, with people using marijuana, cocaine, and other illegal psychoactive drugs; all have had to endure the cultural despoilments and occasionally deliberate depravity of television, rock 'n' roll music, and Hollywood movies; all have grown rich, interdependent, and (in some respects, anyway) strikingly similar since the end of the Second World War. But in one very obvious and tragic way America is an outlier. Its murder rate is one of the highest in the world. [2]

According to the instrumentality theory of American homicide exceptionalism, the explanation cannot possibly be that America has more crime and not more criminals because it does not ("crime is not the problem"). [3] Rather, the finger of blame points, where Zimring and George Newton, and former U.S. Attorney General Ramsey Clark, pointed it thirty years ago: guns. [4] The lethality of interpersonal violence is dictated by the lethality of the instrumentalities at the disposal of antagonists. [5] Guns are hard to get legally in most of the civilized world [Page 970] nowadays: indeed, in Japan, which experiences fewer murders in an average year than does the United States in a few weeks, [6] guns are all but impossible to come by; whereas in America semiautomatic pistols and rifles and powerful shotguns can be had from sporting goods stores with, at most, a few days' waiting. [7] At gun shows and in private transactions, even these minimal restrictions are either inapplicable or are ignored. [8] More than forty percent and perhaps half of American households have guns in them, and in many cases more than one gun. [9] There are far more guns in the United States than there are microwave ovens, [10] automobiles, [11] dogs or cats, [12] nearly as many guns as there are people. [13] Virtually all of these firearms are privately owned. [14] Few are required to be, and even fewer actually are, registered to their owners. If an American is overcome with rage, or with any other urgent reason for wishing to arm himself, chances are he will be able to outfit himself with a gun forthwith. If he does, he will have a means of inflicting wounds that are about an order of magnitude more likely to result in a victim's death than would be knife wounds or blunt trauma. Elsewhere in the rich countries of the world, people are equally as criminally inclined as Americans, as prone to abuse alcohol and each other, but because they cannot readily get their hands on powerful weapons like guns, as Americans can and do, they experience homicide rates only a fraction of those seen here. Or so the argument goes.

The case for this interpretation of experience has seemed so powerful to American opinion leaders that dissenting views have very nearly been swept from public view and relegated to the pages of Guns and Ammo or, occasionally, National Review. [15] [Page 971]

Broadcasters and cable networks have lately been running unanswered public service announcements blaming handguns for the deaths of ten children per day, and there is virtual unanimity among newspapers and mass-circulation magazines that not only is the instrumentality theory correct, but that its correctness is a scientific fact, like Newton's Laws of Motion or Quantum Theory, beyond worthwhile discussion.

There is, however, very deep public acceptance of an interpretation of experience that is in tension with an instrumentality theory of American homicide exceptionalism. Perpetrator theories, as we call them, focus on the motives, intentions, and risk-taking preferences of individuals instead of on the tools ("instrumentalities") that they have selected to carry on their activities. No one should need to be reminded that Americans do believe in human perversity as well as in the connection between perversity and crime. In virtually every part of the country there has been broad political support for large public investments in prisons that no longer advertise rehabilitation as a product they can dependably deliver. Americans are far more supportive of the death penalty than people in other rich countries. Americans are fully prepared to disown public officials who espouse liberal penological philosophies--for example, on the use of prison furloughs or conjugal visitation, as behavioral incentives--and to embrace back-to-the-future, get- tough approaches to prison administration such as chain gangs. These are largely symbolic gestures: what they symbolize is a belief that human perversity exists, and that the proper social response to this perversity is punishment, not therapy.

Though they are not, strictly speaking, logically inconsistent, there are very important conceptual and practical differences between perpetrator theories and instrumentality theories. Theoretically, they see causation in opposite ways: instrumentality theory posits that "gun causes homicide," while perpetrator theory interprets human transactions with lethal outcomes as tending to flow from perpetrators' intentions, either to kill or to run a conscious risk of killing a victim. Instrumentality theory sees guns as an explanation, whereas perpetrator theory sees guns as something that need to be explained (that is, by the recklessness or murderousness of an offender).

More importantly, there is a large practical difference between these two theories. The strong form of instrumentality theory--the sort championed by Zimring and Hawkins--comes [Page 972] close to denying the relevance of criminal justice policy either as an explanation for given rates of homicide or as a tool for changing those rates. Suppose, for example, it is proposed to drive down the homicide rate. Increasing the harshness or certainty of punishment for murder and manslaughter would be barking up the wrong tree if excess undesirable behavior is not seen as a part of the problem. Incapacitative detention (such as "three strikes and you're out" or "truth in sentencing") would be seen as at best a marginal address to the problem unless one were prepared to tolerate much larger prison populations and much more type-I error (that is, falsely supposing the dangerousness of a potential offender). [16] Even changes in the numbers or tactics of police might be thought beside the point. All these interventions aim to change the incentives to commit crimes by application of the straightforward utilitarian technique of altering the likelihood of apprehension or the painfulness of consequences. But if crime is not the problem, such changes might be largely inapposite. And of course in this theory, crime is not ordinarily seen as the problem: "most homicides result from non-criminal social relations"; [17] murderers "are good citizens who [only] kill each other" [18] because a gun was available in a moment of ungovernable fury--" previously law abiding citizens committing impulsive gun-murders while engaged in arguments with family members or acquaintances"; [19] "[t]hat gun in the closet to [Page 973] protect against burglars will most likely be used to shoot a spouse in a moment of rage. . . . The problem is you and me--law-abiding folks;" [20]

Murder is a crime of passion. Few murders are premeditated. The tragedy is that the rifle was standing there in the closet when the son was seized by fury and shot his father, or the pistol was by the bed during a drunken argument and the husband shot his wife. [21]

In this rendering of the problem, the solution, if there is one at all, must revolve around regulating guns: "gun control." Zimring and Hawkins are firmly in this camp, for despite their aversions to the need for "a diversified portfolio of countermeasures to lethal violence," [22] they are as clear currently as in previous animadversions [23] on the subject that gun control is the sine qua non of lowering homicide rates, [24] without which no program for reducing lethal violence can even be taken seriously. [25]

The strong form of perpetrator theory, on the other hand, is completely to the contrary. It holds that, theoretically and practically, instrumentality regulation could have at most a small peripheral value. If a person intends to kill someone, or is willing [Page 974] to risk a killing for the sake of making a credible threat to kill, then chances are he will try to get a gun. Once he is determined to use a gun criminally, making that also a crime is a meaningless, no-value-added gesture; regardless of weapons laws, there will always be enough guns floating around in any modern society to supply the comparatively few persons who want to kill or at least to embrace the risk of killing. As a retired English police official, Colin Greenwood, put the matter:

At first glance it may seem odd, or even perverse, to suggest that statutory controls on the private ownership of firearms are irrelevant to the problem of armed crime, yet that is precisely what the evidence shows. Armed crime and violent crime generally are products of ethnic and social factors unrelated to the availability of any particular type of weapon. The numbers of firearms required [to arm criminals are minute in comparison to the overall number in private hands], and these are supplied no matter what controls are instituted. Controls have had serious effects on legitimate users of firearms, but there is no case either in the history this country [England] or in the experience of other countries, in which controls have been shown to have restricted the flow of weapons to criminals or in any way to have reduced armed crime. [26]

These models imply two quite different regulatory responses. If crime is not the problem, then one should focus efforts where the real problem is--on deadly attacks, rather than on crime. But if deadly attacks are simply a part of an undifferentiated packet of high-risk behaviors that are engaged in by a small and discrete sub-population ("criminals"), then one should expect payoffs, in the form of lower homicide rates and less crime generally, by proceeding in the old-fashioned way, with more intensive and more effective policing and more credible threats of prosecution and imprisonment. It matters which theory is more nearly correct, though both could perfectly well be true (and, to a degree, both certainly are). Even instrumentality theory doubters (for [Page 975] example, the present authors) will affirm that there are at least some "instrumentality homicides." Even the set of most highly-motivated and capable killers would surely find that other weapons could only incompletely substitute for guns; there must surely be some killings by gunshot that would never have resulted if the perpetrator had been armed only with a bludgeon or a knife. But even if one could enumerate instrumentality homicides with certainty (which would disclose whether instrumentality homicides were a large or a small percentage of the total), one would still have to consider the other side of the question--the social utility of private firearms possession. It is only by the trick of assigning that utility some value approaching zero that instrumentality theory generates its accustomed policy prescription.

There are "instrumentality" effects on both sides of the equations that depict human conflict. What mischiefs, including homicides and other serious felonies, were avoided because people with guns prevented them from being completed? Trying to answer this question can all-too-quickly turn into a debate about how one interprets survey research findings about how often civilians use guns to frighten or wound antagonists. Depending upon which estimate one credits, there are currently somewhere between tens of thousands to a few million defensive uses of firearms per year. [27] But the social utility question is hardly exhausted by answering "how many times was a gun brandished (or discharged) in earnest last year?" because of course there must have been n number of cases where no crime was committed, or no attack attempted, because the risk was too great on account of readily available guns for defense. One has to know how much value would be lost if we could move to a world without guns. [28] The question is not at all conjectural, but the data that [Page 976] might serve to answer it certainly are. And, inevitably, discussion of a question so broadened requires consideration of the Second Amendment and its fit into American constitutional institutions.

For the sake of the present discussion, we propose to steer around these very large theoretical and speculative matters insofar as we can. Rather, we shall focus narrowly on the relative plausibility of instrumentality and perpetrator theories of American homicide exceptionalism. Zimring and Hawkins have brought new evidence to the table, which they believe to have iced the cake for the instrumentality theory. We do not agree that it has. It is not the case that murderers are ordinary people (or "good citizens") who had the bad luck to be statistical particles in an America filled up with guns; rather they are extreme and aberrant characters, different from most of the rest of us because of life histories replete with predatory crime, deficient cognitive functioning, and substance abuse. American society controls the most violent left tail of the distribution of its deviant population less effectively than many other countries, but given the vast differences in social cohesion, population characteristics, and the legal and constitutional constraints under which police are required to labor, it is unpersuasive to lay this difference largely at the doorstep of weapons policy.

I. Inferences from Abroad

Summarizing the foreign experience with crime and lethal violence, Zimring and Hawkins write: "If general levels of crime determine the rate of lethal violence, then increases or decreases in crime should result in parallel increases or decreases in lethal violence." [29] Such parallel fluctuation does not, however, occur. Take thirty years of data from the G7 countries. The proposition bears out in the United States, Canada, and Italy (all of which experienced significant increases in theft and homicide between 1960 and 1990), but proves false in Japan (where theft increased, [Page 977] but very little, and an already low murder rate declined substantially) and glaringly false in England, [30] France, and Germany (where theft increased and homicide rates declined or did not change). [31] Can we trust these data? Beyond a certain point it is unnecessary to do so: as Zimring and Hawkins argue, the basic story emerges from the rough cut: property crimes have gone up in all the rich countries since the baby-boom generation began edging into its mid-teens, but murder rates have risen, fallen, or stayed the same.

Now consider the victimization rates (they argue) for various crimes (robbery, assault, all violent offenses, and property offenses) in twelve to eighteen countries including the United States. For property crimes between 1988 and 1991, one finds the victimization experience of the United States at the high end of the distribution, essentially tied with such countries as Spain, Canada, the Netherlands, New Zealand, and Australia. [32] For "all violent offenses" during the same period, the United States is again at the high end of the distribution, tied with Canada, New Zealand, and Poland and significantly behind Australia. [33] For robbery rates (with 1991 as the for instance), again the United States victimization numbers are at the high end, along with Italy, Australia, England, Canada, and Czechoslovakia but well behind Poland. [34] For assault, the United States is distinctly a second-class power. Canada, Australia, and New Zealand lead the league; America is back in the peloton tied with Finland, Poland, and the Netherlands. [35]

These numbers depict a fairly straightforward story. The United States, like other Anglophone countries, is a pretty tough town, near but not all the way at the head of the class for most categories of crime. It is not, as Zimring and Hawkins point out, very different-looking from comparison countries until one looks at lethal outcomes measures. "What separates the United States from the rest of the industrial Western world . . . is [its] relatively [Page 978] small number of life-threatening attacks, usually shootings. . . ." [36] Here, America is in a class by itself.

It remains to be established, however, what all this information means. The question begged by these data, the implication that we are plainly invited to draw from them, is that a priori one should expect a certain uniformity in the relationship between crime rates to lethality rates when comparing the United States and other countries. But this implication lies well beyond either logic or experience. Perhaps American robbers and assaulters are simply a deadlier lot than Australians, Frenchmen, Englishmen (and so on), better armed because they are prepared to be more violent than their foreign counterparts. Though plenty of evidence can be adduced to the effect that the United States is rather like other industrialized countries, there is plenty of other evidence that in fact the United States is rather unlike other industrialized countries. Consider, for example, the case of rape--a crime that seldom involves the use of a gun. [37] The United States has far higher rates of rape than other industrialized countries: for example, it is anywhere from thirteen to twenty-three times higher than the English rate; [38] almost five times the French rate; [39] almost six times the German rate; [40] and thirty times the Japanese rate. [41] All of these measurements approximate the homicide rate disparities that show up when the United States is compared with these other countries. [Page 979]

It would be a treat to hear an instrumentality-based explanation for American rape exceptionalism, but the point that is illustrated by this disparity is too important to be let stand as a joky debater's taunt. There are large differences between Americans and the populations of foreign countries, even English-speaking countries with many superficial similarities. American rates of non-firearms homicide are far higher than homicide rates from all causes in England and Japan, [42] a circumstance that supports the inference that murder rates in the United States would be many times higher than English or Japanese rates even if every civilian firearm in this country suddenly vanished in a puff of smoke. The United States records more non- firearms murders than there are murders from all causes in Western Europe and Japan combined, territories more than half again as populous as the United States.

The culture here, at least in some crucial aspects, is quite different from anything found in other rich countries of the world. These differences far transcend the question of firearms and how they are distributed. They dictate how young men will answer narcissistic injuries and territorial affronts, and what one may or may not do in order to get money or drugs or sex (and so on). One should hardly be surprised to see big differences from one place to another in what sorts of behavior will get one arrested, how various categories of crime are defined, when violence, including killing, is most to be expected, and whom we should expect to be involved. American criminals and those of England, France, and so on, cannot usefully be compared directly without a far more detailed account than is presently available concerning their similarities and differences. Without such an account, the instrumentality theory of American homicide exceptionalism is reduced to a tendentious assumption. Demand for firearms on account of their lethality could be all one could ask for to explain the unusual prevalence of firearms homicide in the United States.

One very obvious difference between the people of the United States and those of other rich countries is our thirty million or so African descendants. African Americans have, and long have had, rates of homicide that are eight to ten times higher than those of European- or Asian-American populations, both as [Page 980] perpetrators and victims. So while African Americans amount only to twelve percent of the United States population, [43] they account for more than half of the homicides. [44] The reasons for these disparities are not fully understood, but the existence of the phenomenon is undeniable. Among some of the more plausible explanations for this picture are: that crime is an activity largely confined to young males, and the African-American population is younger on average than that of the United States as a whole; [45] the absence of African Americans--for long by systematic, legal exclusion--from the "institutions of urban-industrial discipline and restraint," [46] such as factory and other unionized jobs; that African-American urban sociology reflects what anthropologist Elijah Anderson has called "the Code of the Streets"--" the issue of respect[,] . . . [being] granted the deference one deserves" [47]--a besetting problem in any group of any size where norms are generated and sustained by unmarried young men with weak future prospects; [48] human capital theory (and its cousin "racism") which would predict an inverse correlation between wealth and risk-taking (including criminal) behavior; [49] and the weakness of African-American families, recognizing that families are the institutions in which boys are--if they ever are--taught forbearance and restraint. [50]

Another explanation for why there are so many more extremely violent criminals in the United States than in Europe was put forward by the eminent historian C. Vann Woodward:

The influence of European immigration on American history has received much attention. But the impact upon Europe itself of the emigration of 35,000,000 Europeans in the [Page 981] century between the Napoleonic Wars and World War I remains to be acknowledged . . . . The importance of the [American] West as a safety valve for American society has undoubtedly been exaggerated. But the significance of America as a safety valve for Europe and the effect of the closing of that valve after World War I remain to be fully assessed. [51]

It is beyond the scope of the present essay to propose a grand explanatory synthesis for racial (or other group) differences in rates of crime or violence, and sufficient simply to note that the differences are real, large, and persistent. When it comes to violent crime, the different subcultures and subpopulations found in the United States have rather different experiences from one another, and rather different experiences from people elsewhere in the world. So much is obvious, not very interesting, and not at all of comfort to any instrumentality theory of American exceptionalism. "History" and "culture" name everything and therefore explain nothing, at least not without a great deal of particularizing and explication. We do not pretend that we can deliver a rounded theoretical alternative to the instrumentality theory to explain national differences in homicide rates (or national differences generally). But it does not really "take[] a theory to beat a theory" [52] if the theory to be beat is just obviously wrong, [53] at least in its stronger forms, and so riddled with aberration that it cannot possibly be correct. Guns do kill people, of course, but they need plenty of help from people with lethal motives in order have this effect. Portraying the phenomenon of American homicide exceptionalism as a function of our firearms diffusion statistics is a narrative tour de force without constructive implications for public policy. [Page 982]

II. More Guns, Not More Killing, I

There is a poor correlation between firearms numerosity and diffusion and homicide rates. This observation appears in several guises. Gary Kleck reports, on the basis of as-yet unpublished data, [54] that homicide and gun homicide compared to (legal) gun ownership figures from thirty-six nations shows no correlation: lower rates of legal firearm ownership did not coincide with lower rates of homicide or gun homicide; neither did higher legal ownership rates coincide with higher rates of homicide. [55] The same null correlation appears, incidentally, from a time series study based on records of (legal) English firearm ownership and crime from 1870 to date: lower rates of (legal) firearm ownership did not coincide with lower rates of homicide or other violent crime; neither did higher legal ownership rates coincide with higher crime. [56] If these findings are right, the instrumentality model is wrong.

Consistently in England, Canada, and the United States, areas with the greatest legal gun ownership have the lowest rates of violent crime compared to other areas of the same nation having lower gun ownership. [57] This is "a curious fact if firearms stimulate aggression" (as some versions of instrumentality theory [Page 983] would suggest), according to criminologists Hans Toch and Alan Lizotte. [58] Addressing the issue as to American statistics alone, they conclude "that guns do not elicit aggression in any meaningful way." [59] "Quite the contrary, these findings suggest that high saturations of guns in places, or something correlated with that condition, inhibit illegal aggression." [60] By curbstone empiricism it looks as though something similar is reflected in state-by-state crime reports. Jurisdictions where household firearms are reputedly very common-- Northern New England, the Dakotas, Idaho, and Wyoming--have the lowest homicide rates in North America--quite European homicide rates, in fact--along with sub- European crime rates generally. Crime is not their problem; neither is homicide. [61] Contrariwise, very high homicide rates appear to be consistent with low diffusion rates similar to those observed in a number of European countries with fairly low homicide rates. [62]

[Page 984]

Table 1: Murder Rates in Relation to Firearms Inventory, 1975-1996 [63]
Year	Total Gun Stock		Guns per 1,000 pop.		Homicide Rate per 100,000
	Handguns	All Guns	Handguns	All Guns
1975	41,081,714	139,915,125	192.8	657.1	9.6
1994	84,565,690	235,504,001	324.8	904.6	9.0
Year	Extrapolated Gun Stock				Homicide Rate per 100,000
	Handguns	All Guns
1995	87,973,867	241,806,685			8.2
1996	91,382,044	248,806,685			7.3 (preliminary)[64]

It is also the case that American firearms stocks, especially handgun stocks, have risen very sharply as the national homicide rate has markedly declined. [65] Such a remarkably nonlinear "dose-response" relationship is more reasonably interpreted as evidence of noncausality than the converse. Instrumentality theory predicts that there should be at least a loose connection between the stock of firearms and the homicide rate such that a massive increase in the former should lead to an obvious increase [Page 985] in the latter. Nothing like this seems to have happened. Civilian ownership of all kinds of firearms increased by 68.3% between 1975 and 1994, and handguns increased by 105.8%. [66] The murder rate was falling through most of this period actually decreasing by 6.3%. [67] The year 1995 saw the addition of 6,302,685 more firearms, [68] yet by its end homicide had declined 14.6% from its 1975 level. [69]

[Page 986]

The homicide rate was disobligingly erratic over this twenty-year period, peaking in 1979, falling through the early 1980s, jumping up toward the end of that decade through the early 1990s, then falling sharply back again. [71] Nobody has a full understanding of what drives these changes any more than what drives changes in hemlines or television programming, but "firearms" begin to look more like usual suspects than plausible ones given that, while homicide rates in the 1975 to 1994 period "present[ed] a picture of oscillation around a strikingly flat trend," [72] firearms stock or handgun stock moved smartly in only one direction over that period of time. When the murder rate peaked around 1980, handgun ownership stood at 51.7 million; as of the end of 1995, there were at least 36.2 million more handguns while the homicide rate was down by twenty percent.

At the very least, then, it cannot be denied that very large increases in civilian-owned firearms and very large decreases in the homicide rate are compatible phenomena. What is even more embarrassing to an instrumentality account of the world, however, is the kind of killings that have increased and declined since 1970, when Ramsey Clark ascribed murder to the rifle "standing there in the closet when the son was seized by fury and shot his father, or the pistol [that] was by the bed during a drunken argument and the husband shot his wife." [73] Homicide within the family has markedly declined since then. [74] The use of firearms in such homicides is lower (54.7%) than in felony-type murders (65.7%) and gangland and juvenile gang slayings (88.4%). [75] It is not that domestic murderers are a fundamentally [Page 987] more peaceable lot than other killers; they are at least as likely to have violent histories as those whose victims were not members of the same household. Domestic murderers use guns less, in all probability, because they are less needed in the home than when attempting to coerce, for example, a rival gang member or a recalcitrant customer. But of course over the past generation, guns of many sorts, handguns in particular, have proliferated in people's homes.

American homicide statistics present an arresting puzzle, though hardly the one that instrumentality theorists have identified: why have Americans been getting more peaceable for the past twenty years? Murder rates among acquaintances have been falling. Not only does one see acquaintance homicides as a subsiding fraction of all homicides; to judge from Chicago's numbers they seem to be, even taking account of the city's shrinking population, declining absolutely. Indeed, if one factors out the unprecedented--and undoubtedly cocaine-linked [77]-[Page 988] upsurge of minority teenage violence that inundated the crime statistics between 1984 and 1992, one sees American homicide rates falling steadily across the boards.

This, then, is the criminological puzzle of the decade: why are Americans becoming so much more peaceable? This question is exactly opposite to the one that conventional wisdom suggests, and what is more, the pacification of America is occurring in the presence of a large and ever-growing arsenal of handguns.

III. More Guns, Not More Killing, II

The largest econometric study ever done of firearms and crime fails to identify any obvious or regular connection between relaxing laws that regulate the concealed carrying of handguns and increases in rates of homicide. [78] Indeed, that famous study, by John Lott and David Mustard (the "Lott-Mustard paper"), purported to find just the opposite: large effects on the kind and quantity of crimes linked to a law reform whose purpose was to make the carrying of handguns far more common than it had previously been. The transition to liberalized concealed handgun carriage was vigorously debated in every jurisdiction that adopted it. There is extensive evidence, in the form of newspaper editorials and public statements by opponents of the reform, that previous work on gun control by Zimring and Hawkins had found a wide audience. [79] Everywhere the cry was the same: let a few tens or hundreds of thousands of people carry guns around with them and the streets would run red with blood; [80] Texas [81] (or Florida, [82] or Virginia, [83] or North Carolina, [84] or wherever), would [Page 989] become the new Dodge City, the object of ridicule of every civilized place, as every fender bender, football argument, and romantic quarrel turned into a deadly confrontation. Obviously these laws would make it more likely that potential antagonists would be carrying guns than had been true before. What is worse, every armed person who was drawn into a quarrel with a furious stranger would be entitled to surmise that their antagonist was also, more probably than before the law was changed, armed with a gun and a motive to use it. [85] It is an unstable situation, to be sure, when self-preservation counsels one to shoot first and asks questions afterward. Critics of this reform have invariably predicted that this circumstantial instability must lead to unlimited quantities of bloodshed. [86]

There is, moreover, plenty of territory in which to test this prediction: it need not be left as a sterile argument between untestable (or untested) models of reality. Over the past decade or so, thirty-one states have liberalized laws regulating the carrying of concealed weapons. Under these laws applicants are entitled to be issued a license allowing them to carry a concealed handgun on their person upon proof that they are law-abiding, responsible adults who have firearms training. The number of applicants taking advantage of the laws to obtain concealed carry permits varies from state to state, and although sometimes significantly higher, is usually one or two percent of the eligible population. [87] All told there appear to have been more than a [Page 990] million concealed carry licenses issued, with licensees in large population states like Connecticut, Florida, Pennsylvania, and Texas numbering in the hundreds of thousands. [88]

A funny thing happened on the way to the OK Corral, however. For some reason the predicted blood-bath from liberalizing concealed carry laws has not come to pass. No wrongful [Page 991] homicides have been reported, in any jurisdiction, while such licensees were carrying their handguns. Irregular incidents of any kind have been few and far between. Police have reported no problems resulting from the issuance of a million plus concealed carry licenses, [89] and legislators, including the leading opponent of the law's enactment in Florida, and other officials who strongly opposed the concealed carry laws have had to admit their fears have not been realized. [90] "'I'm eating a lot of crow on this issue' [said Houston area prosecutor John] Holmes, a vocal opponent of [Texas' enactment of] the law." [91] "I have changed my opinion," says John Dunn, Sheriff of Campbell County, Kentucky. [92] "I am glad that I was wrong," says a North Carolinian police chief who had also opposed the law, "I haven't had a single report of a licensed concealed weapon being used to commit a crime." [93]

The more spectacular claims of the Lott-Mustard paper (that is, that increasing concealed carry of handguns has a large suppressive effect on violent crimes) are certainly controversial and may well turn out to be premature. [94] But whether guns are good medicine for crime is not the same question as whether they [Page 992] are bad medicine for the homicide rate. They are not bad medicine for the homicide rate. The thirty-one-state trial of liberalized concealed carry laws, including about half the United States population, has set up a very large-scale natural experiment that tests the hypothesis that more guns lead to more killing. No one has read the experience thus far as consistent with the "Dodge City" prediction--anywhere, let alone everywhere.

To believers in the instrumentality theory, this should seem a decidedly curious result. There are two, and only two, possible explanations for it. One of them is that liberalizing concealed carry laws actually does not lead to appreciable changes in patterns of gun-carrying behavior (nor to changes in people's perceptions about how gun-carrying behavior had increased). [95] This explanation, if true, deprives adversaries of such laws of their best reason to oppose them. The other possibility is that the instrumentality theory ("more guns, more killing") is false. We suggest that the second interpretation is much the more plausible.

IV. The Character of Murderers

A crucial link in the instrumentality story of American homicide exceptionalism is the proposition that firearms availability turns ordinary citizens into circumstantial candidates for the state penitentiary: "[T]he majority of homicide victims die not as a result of criminal activity, but because of arguments between people who know each other"; [96] "[T]he social processes that generate arguments that result in homicides are not distinctively criminal in most cases"; [97] "[They arise] out of arguments and [Page 993] other social encounters between acquaintances"; [98] "[M]ost homicides result from noncriminal social relations"; [99] "Most of the circumstances that generate homicide are not property crimes involving strangers, but arguments among acquaintances that nobody would regard as distinctively criminal until the attack began." [100]

Similarly, most murderers are "neither felons nor crazy," but just "people involved in family fights and fights over jobs and money, and people who are sad or depressed" [101]--" previously law abiding citizens committing impulsive gun-murders while engaged in arguments with family members or acquaintances." [102]

There can be little question that criminals equipped with guns are more dangerous than those without. But it does not follow, and in fact it is not true, that law-abiding people who possess guns are likewise more socially dangerous. There has now accumulated a good deal of direct evidence concerning the identity and background of murderers, that makes it evident that the "plain folks" account can no longer be wrapped in the mantle of science. [103] Murderers are not plain folks. Neither, for the most part, are their victims. To the extent that killings are the product of arguments, it should be borne in mind that only arguments between a certain sort of antagonist are likely to have a mortal outcome.

It has long been clear that murderers are, in comparison with the general population, disproportionately likely to have significantly [Page 994] subaverage intelligence, suffer from insanity or brain damage, alcoholism or substance abuse, or combinations of these problems. Numerous American, Canadian, and European studies find that people with impaired cognitive capacity, mental illness, or a history of substance abuse are several times more likely to engage in violent crime in general than are ordinary citizens. [104] Murderers exhibit this same predisposition:

Although there is some controversy [about the distinctiveness of murderers from other criminals], it has long been suspected that homicidal behavior might result from brain dysfunction associated with impaired ability to regulate goal-directed activity and modulate impulsive response. These neurological risk factors include abnormalities in brain functioning, pre- and perinatal influences, history of head injury, delayed psychomotor and speech development, and learning difficulties. [105]

An early onset of alcohol abuse is one of the hallmarks of Type II alcoholism, a condition that has been related to the propensity for impulsive, violent crimes under intoxication. Substance abuse also may be a part of an adult lifestyle among violent and impulsive offenders. [106]

[Among violent and nonviolent Finnish criminals] the homicidal group tended to differ from the other groups in cruelty to animals during childhood, had an earlier onset of alcohol abuse [than either of the other groups] and a higher prevalence of drug dependence than those in the nonviolent groups, and had experienced over three times more physical [Page 995] abuse than offenders in the nonviolent group. . . . These results are congruent with earlier research findings . . . . [107] An American review of homicide studies notes that the "research literature on characteristics of those who murder yields a profile of offenders that indicates many have histories of committing personal violence in childhood against other children, siblings, and small animals." [108]

From the 1960s to date, national and local studies document that seventy to eighty percent of murderers have formal adult criminal records. [109] Prior to an uptick in the mid 1980s, minors--who practically never have adult records--had committed ten percent of murders; more recently minors' "market share" has been over fifteen percent. [110] Of those killers who are old enough to have adult records, ninety percent or more do. Most of these criminals are not even thirty years old, but in their few adult years have averaged four major felony arrests over a six-year crime career. [111] (By contrast, something like eighty-five percent of the population has never been arrested even once). [112] [Page 996]

Based on a national sample of prison interviews, the Bureau of Justice Statistics estimates that in the 1980s, about 13,200 murders were committed while the offender was on parole or probationary release for a prior felony conviction. [113] At the time they committed murder, eleven percent of killers had been released on bail or recognizance while awaiting trial for some previous offense. [114]

These formal criminal records, moreover, represent only the proverbial tip of the iceberg. Many of these individuals have prior crime histories which are unknown because either they were not caught [115] or their crime record was as a juvenile and hence is confidential and, perhaps in addition, expunged. [116] One of the few studies to obtain access to juvenile records, that of David Kennedy and his colleagues at Harvard's Kennedy School, analyzed Boston-area juvenile gun and knife murders for the years 1990 to 1994. In their less than twenty-one years of life, the 125 killers in this sample had had an average 9.7 prior arrests--three arrests for prior murders, hundreds for prior gun crimes and non-gun violent felonies, and even more for drug and property crimes. [117] It is a safe bet that many adult murderers' juvenile records would supplement and extend their average of four adult major felony arrests. It is not a mistake of the ordinary sort but rather an extravagant falsification of reality to claim, as some researchers have, that such individuals "would be considered law-abiding citizens prior to their pulling the trigger." [118]

Furthermore, not even the small fraction of murderers without a prior criminal record--ten percent or so of the total--can accurately be characterized as "law-abiding citizens." A few fit that characterization, but many will have come to the attention of the police repeatedly for matters, such as domestic battery, [Page 997] in which, as a matter of custom, arrests have been sparingly made. Studies show "most family murders are prefaced by [the killer's] long history of assaults"; [119] in ninety percent of cases, the police had been called to the same address at least once within the preceding two years; the median number of prior police calls to that address was five during that period. [120] But none of those five or more disturbance calls may have ended in an arrest. True, perhaps as many as seventy to seventy-five percent of domestic homicide offenders have previously been arrested. [121] Significantly, however, domestic murderers may have a long history of brutalizing within the family yet only have a record for attacking others outside the family. Massachusetts data show that seventy-five percent of adult wife-batterers have a prior crime record, but half their prior violence arrests were for attacks on nonfamily members. [122]

"[W]ith comparatively few exceptions, homicide reflects a long-standing pattern [of the murderers' prior similar] behavior." [123] Because homicide, even among the most "at risk" populations, remains relatively rare, one cannot reverse the assertion that "most killers were criminals" and pretend to know which individuals, among the population of likely candidates, actually will go on to commit murder. [124] But murderers remain a highly unusual and aberrant population. Recognizing this is the key to understanding why arguments between murderers have a relatively high risk of deadly issue, but arguments between "plain folks," even well-armed plain folks, almost never have a lethal outcome.

At least with respect to their criminal history, it is evident that victims more resemble their killers than they do the population generally. David Kennedy found that Boston-area juveniles who murdered other juveniles averaged 9.7 prior criminal arraignments in Massachusetts courts, compared with their [Page 998] victims' average of 9.5 criminal arraignments. [125] "Probation officers said it was rare to lose a kid to gun violence who had not previously been on probation, and the dark joke among gang officers was whether it would be ethical to [get rich by] tak[ing] out life insurance on certain kids." [126] At least sixty percent of those homicides were by or among gang members. "This is a conservative estimate, since [it did not include] cases in which the circumstances were unknown or ambiguous . . . ." [127] Philip Cook and John Laub found that, nationally, guns are used in eighty-eight percent of adult and juvenile gang murders. [128] Typical of findings in local studies across the nation are that seventy-one percent of minors wounded in Los Angeles drive-by shootings "were documented members of violent street gangs"; [129] that eighty percent of homicides in Washington, D.C., are "drug-related"; that nearly seventy percent of admittees in 1994 through 1996 at a major Maryland trauma center who had been shot, knifed, or beaten tested positive for alcohol, cocaine, opiates, marijuana, amphetamines, or phencyclidine and many were substance-dependent; [130] and that eighty- four percent of 1990 homicide victims in Philadelphia had crime records or showed antemortem use of illegal substances (or both). In addition, the Rand Foundation study, "Homicide in Harlem," estimated that "the great majority of both perpetrators and victims of assaults and murders had previous arrests, probably over eighty percent or more." [131] A study of gunshot wounds reported to the Charlotte, North Carolina, Police Department in 1992 and 1993 found that seventy-one percent of adult victims had criminal records. In Chicago, the Police Detective Bureau reports that nearly two-[Page 999] thirds of victims (as compared with three-quarters of apprehended perpetrators) have adult arrest records. [132]

The murder wave that began in the mid 1980s affected only a very narrow band of the American population. Through most of that decade, nearly all of the country continued to decompress from the baby-boomer crime wave that peaked before 1980. Rates of victimization by persons over age twenty-five, regardless of race, declined more or less continuously from 1980 onward. [133] But between 1985 and 1990, the murder rate among young African- American males increased extremely. The victimization rate for black males in their middle and late teens nearly tripled in those years, from 37.4 in 100,000 to 105.3 in 100,000. For men between ages twenty and twenty-four, it more than doubled (from 63.1 in 100,000 to 140.7 in 100,000). The victimization rate for the population as a whole is less than 10 in 100,000.

The crucial thing is not that many murders arise out of arguments, but who is doing the arguing. Everyone experiences interpersonal conflict in one way or another and at one time or another, but it is exceptionally rare for such antagonisms to lead to mortal consequences unless at least one of the disputants has a life history of crime (often irrationally violent), mental disorder, or substance abuse. The mystery of why, contrary to all common experience, arguments lead to murder disappears when we realize who the participants in these particular arguments are: drug dealers versus customers or competitors; co-members of criminal gangs or rival gangs; and abusive men versus the women and children they have brutalized on countless prior occasions.

IV. International Comparisons

Two distinct issues are broached by the instrumentality theory of American homicide exceptionalism. One is how the distribution of firearms is related to crime, murder, and so on. The other is how regulatory laws are related to the distribution of firearms. For many years, confusion between these two propositions has been a stable feature of firearms policy discourse, [134] but [Page 1000] the two questions are quite distinct. People's persistent refusal to obey, for example, anti-pornography laws, anti-narcotics laws, or highway speed limits (especially in wide-open, empty places like the Dakotas), should serve as a reminder that passing a law is one thing and getting people to obey it is another. Indeed, compliance with any law will certainly be inflected by a number of variables.

First and most conventionally, people calculate the probability of being apprehended if they disobey a law and the severity of the punishment if they are apprehended. This is usually called the general deterrence principle: severity of punishment discounted by probability of apprehension. The general deterrence principle predicts that, other things equal, one should see less and less compliance with any given law as it becomes more and more difficult for authorities to detect violations or as it becomes more and more difficult to persuade juries to punish people who are caught. But a number of other factors must certainly be as important, if not more important, to compliance. For example, people cannot be expected to obey a law they do not know about. [135] Then there is the matter of institutional respect: other things equal, one would expect less obedience of a law among people who did not respect the jurisdiction of the enacting authority than among people who did. [136] Convenience is often important: one should expect to see more obedience to a law where it is easy to obey than where it is difficult. [137] People certainly calculate how obedience to a law will likely impinge on their own private welfare: heroin addicts, fearful of withdrawal, are less likely to obey narcotics laws than otherwise similar people who have never touched heroin. Finally, a "self-reputation" for law-abidingness is important to most individuals. [138] People engaged in arrantly illegal behavior commonly tell themselves face-saving, self-justifying stories (it's not really illegal; everybody does it; I have a right; I am only reclaiming a patrimony which was stolen, by the capitalists or the Elders of Zion or whoever). [Page 1001]

The moral of this story is that one must suppose that compliance with prohibitory gun control laws (for illustration, a national law banning handguns) would be uneven in any heterogeneous population. One would expect the most rapid compliance by handgun owners who did not think they actually needed a handgun for anything and who respected the authority of the national government to prohibit such weapons. One would expect relatively lower levels of compliance among people who were convinced they needed a strong form of self-defense (especially people who, rightly or wrongly, distrusted the police) and who were either outright contemptuous of law or adroit at making up self-justifications for noncompliance with it. The second category by definition includes more criminals and sociopaths than the first. Assuming compliance with gun control laws does, over some period of time, lead to appreciably fewer weapons being in circulation, it would seem to be obvious-- axiomatic--that throughout this period of transition, criminals and sociopaths must be increasingly well-armed, in relation to the population as a whole--and even more so in relation to the most law-abiding tail of the distribution. The instrumentality theory must offer some sort of justification for believing that the relationship between firearms distribution and crime, lethal outcomes, or any other variable output will approximate a monotonic function over a twenty- or thirty-year period of time, so that in each successive period in which one observes fewer firearms diffused among the population, one also observes, or tends to observe, less homicide, suicide, and other serious violence.

There is no such direct evidence, but indirect evidence might arguably be derived by comparing the experience of the United States and that of other countries. Compare the (relatively) high homicide (or suicide) rate in Seattle, Washington (where handgun control is lax), and the relatively low rates in Vancouver, B.C., (where it is strict): res ipsa loquitur. [139] Compare London with New York and see how, with similar rates of burglary and assault, gun-happy New York outstrips gun-controlled London in terms of lethal outcomes. [140] Do not such cross-sectional comparisons [Page 1002] offer strong evidence that gun control policies actually do work?

It is a standard move in instrumentality theory argumentation to offer international or cross-cultural comparisons to establish the proposition that American homicide exceptionalism is essentially explained by American gun laws or the lack of them. [141] But it is far from obvious what one can learn from this technique without controlling for relevant differences in the populations being compared. After all, differences in gun control laws--and, for that matter, differences in the actual distribution of firearms, which is a different and surely far more important matter--are seldom going to be the main differences between a pair of cities each in a different country. Without adequately accounting for other variables, most obviously, differences in their demographic makeup, these sorts of comparisons run the risk of degenerating into mere tendentiousness.

Indeed, one indication of the loose fit between gun control laws and homicide rates can be drawn from historical experience. American homicide exceptionalism long antedated modern gun control laws. Years before European countries adopted the restrictive firearms policies that most of them have today, [142] they [Page 1003] enjoyed very low rates of violent crime. [143] In the absence of prohibitory legislation, England and continental Europe were important markets for so-called pocket pistols (small, light- caliber handguns), which were fabricated in innumerable small factories all over Europe, [144] and in the years preceding the First World War, at least a million .32 caliber Browning pocket pistols, and a hundred-thousand .25 caliber pistols, were shipped to Europe from the United States. [145] These countries' gun control laws were nothing like as strict as American laws during this period, [146] though American rates of criminal violence, then as now, were far higher than European rates. It is an anachronism to locate American homicide exceptionalism primarily in national differences in firearms regulation. As historian Eric Monkkonen has pointed out:

Virtually every analysis put forward to explain the [comparatively] very high United States homicide rate has been [Page 1004] ahistorical. . . . Had they been proposed as historical, they would have foundered quickly for the explanatory inadequacy of these "pet" theories becomes immediately apparent in a historical context. [147]

It is, moreover, an historical solecism to hold the English nation up as an exemplar of peaceable law-abidingness or, for that matter, of the effectiveness of gun control laws as a means of controlling homicide and other crime. During the Middle Ages, in fact, England without guns may have had a murder rate more than double the current American murder rate. [148] The "English were noted throughout Europe for their turbulence and proclivity to violence." [149] It was not until the nineteenth century that England began to assume its modern reputation for peacefulness, and by the time it adopted restrictive gun control legislation, the Firearms Act of 1920, [150] England ranked among the world's most placid nations. Yet prior to 1920, the principal gun control law in the country was that policemen would not carry firearms.

English and European gun control originated as a reaction, not to crime, but to the political violence which racked the continent after the First World War. [151] Weimar Germany adopted similar laws for similar reasons about the same time. [152] The laws were enacted in order to cope with the violence, yet (although one [Page 1005] cannot, of course, be certain) it seems fair to observe that statutory gun control measures did not obviously have much of a pacifying effect either on crime or political violence. The United States and Switzerland (another country where household firearms have always been encouraged) have suffered far less from assassination, terrorism, and political violence than have England or continental Europe through most of the twentieth century.

Moreover, England and Europe have seen ordinary apolitical violent crime increase over the past seventy years, to as great or even a greater extent than has the United States (although of course from much lower base levels). [153] "No matter how one approaches the figures," Colin Greenwood wrote in the early 1970s, "one is forced to the rather startling conclusion that the use of firearms in crime" in England "was very much less" before 1920 "when there were no controls of any sort and when anyone, convicted criminal or lunatic, could buy any type of firearm without restriction. Half a century of strict controls has ended, perversely, with a far greater use of this class of weapon in crime than ever before." [154] Since then, violent crime, and particularly firearms crime, has so steadily increased that ever increasing numbers of English police are being armed. [155]

Greenwood is not to be read as arguing that it is advisable to allow lunatics and criminals to purchase firearms. The point is, rather, that relatively little irregular behavior with firearms can be attributed to people with innocuous life histories, whereas those from whom one has most to fear, criminals and lunatics, have relatively little compunction about defying or evading laws, gun control laws among them, meant to control their behavior. [Page 1006]

V. Implications for Public Policy

The connection between guns and American homicide exceptionalism becomes explicit only in the population of criminals. That is why crime is after all the problem, and why changing the subject to homicidal violence, on the pretext that killing is a stochastic property of armed populations that are not in other respects predisposed to violence, is seriously misleading. The challenge for public policy is not how to further burden the lives of the innocuous many but rather how to manage the antisocial behavior of the toxic few. What one can learn from the experience of other countries that would help attain this objective is pervasively questionable. Obviously foreign countries will have quite different rules of police practice and criminal procedure than those applicable in the United States; both as a formal legal (or constitutional) matter and as a matter of sociology, authorities are differently constrained, and undoubtedly in most cases less constrained, than their United States counterparts. It follows that the crime-control (or violence-control) lessons one could possibly learn by studying (for example) Japan or Italy are going to be of limited applicability in the United States unless we are able also to import the people, customs, traditions, criminal procedures, shared habits of mind, and social organization of Japan or Italy. One would probably do better, in the long run, presuming that institutions cannot be transplanted from one culture to another than presuming that they can. [156]

That is not to deny that it makes common sense to prohibit people with histories of violence or crime from possessing guns. Even though it is difficult to catch someone in unlawful possession of guns (or anything else small enough to conceal), no one will deny that police ought to arrest such a person if, perchance, they should encounter one. But that is not a strategy that suggests trying to limit access to firearms by the risky few through controlling the behavior of the harmless many, whose possession of firearms raises their murder risk not at all in [Page 1007] statistical terms, and their suicide risk, at most, very little. [157] There is after all,

no evidence anywhere to show that reducing the availability of firearms in general likewise reduces their availability to persons with criminal intent, or that persons with criminal intent would not be able to arm themselves under any set of general restrictions on firearms. [158]

Perhaps one can agree, in the end, that it is not crime, but criminals that really are the problem. But this is a semantic quibble that leaves the conversation precisely where it was before Zimring and Hawkins's latest contribution.

* Kirkland & Ellis Professor of Law, Northwestern University; J.D., University of Minnesota.

** Partner in the firm of Benenson & Kates; LL.B., Yale University.

[1]. See Franklin E. Zimring & Gordon Hawkins, Crime Is Not the Problem: Lethal Violence in America 27-33 (1997).

[2]. See id. at 8 fig.1.3.

[3]. Id.; see also Franklin E. Zimring & Gordon Hawkins, The Citizen's Guide to Gun Control 13-20 (1987) [hereinafter Citizen's Guide].

[4]. See Citizen's Guide, supra note 3.

[5]. See id. at 15-16.

[6]. See Zimring & Hawkins, supra note 1, at 7-8.

[7]. See, e.g., Colo. Rev. Stat. § 12-26.5-103 (requiring only an instant criminal background check on the firearms purchaser); Cal. Penal Code § 12076 (West 1991).

[8]. See Melvin Claxton & William Gaines, Fire Sale: America's Unchecked Gun Market, Chi. Trib., Dec. 29, 1997, at A1.

[9]. See Gary Kleck, Point Blank: Guns and Violence in America 17-64 (1991).

[10]. See Bureau of the Census, U.S. Dep't of Com., Statistical Abstract of the United States 1997: National Data Book 728 tbl.1207 (1997).

[11]. See id. at 627 tbl.1009.

[12]. See id. at 255 tbl.412.

[13]. See id. at 28 tbl.26.

[14]. See Gary Kleck, Targeting Guns: Firearms and Their Control 96- 97tbl.3.1 (1997).

[15]. See, e.g., Daniel Polsby, The Closed American Mind: From the Hip, Nat'l Rev., Mar. 24, 1997, at 33.

[16]. See Franklin E. Zimring & Gordon Hawkins, Incapacitation: Penal Confinement and the Restraint of Crime (1995).

[17]. Zimring & Hawkins, supra note 1, at 61. The quoted language is taken from a rhetorical question, which the authors answer thus: "The social processes that generate arguments that result in homicides are not distinctively criminal in most cases." Id. Compare the following self- description given by a group devoted solely to anti-gun advocacy, the Violence Policy Center:

[T]he majority of homicide victims die not as a result of criminal activity, but because of arguments between people who know each other. The reality of firearms violence is that it stems not from "guns in the wrong hands," but from the virtually unregulated distribution of an inherently dangerous consumer product ....

Violence Policy Center, About the Violence Policy Center (n.d.).

[18]. National Coalition to Ban Handguns, A Shooting Gallery Called America (n.d.) [hereinafter National Coalition].

[19]. John Lindsay, The Case for Federal Firearms Control 22 (1973). A Shooting Gallery Called America cites The Case for Federal Firearms Control in further asserting that "each year" thousands of "gun murders [are] done by law- abiding citizens who might have stayed law-abiding if they had not possessed firearms." National Coalition, supra note 18 (emphasis in original); and "that most murders are committed by previously law abiding citizens where the killer and the victim are related or acquainted." Id.

[20]. David Kairys, A Carnage in the Name of Freedom, Phila. Inquirer, Sept. 12, 1988 (emphasis added).

[21]. Ramsey Clark, Crime in America 108 (1970); see also Editorial, Run Down, Gunned Down, Wash. Post, Mar. 4, 1994, at A22; George Edwards, Murder and Gun Control, 18 Wayne L. Rev. 1335 (1972). In this regard, Dee Helfgott, the coordinator of Coalition for Handgun Control of Southern California, Inc., stated:

Most murders (seventy-three percent) are committed impulsively by previously law abiding citizens during arguments with family members or acquaintances.... A readily available gun is what turns an assault into a murder. The handgun, often kept in the home for self-defense, is six times as likely to be used against a family member as it is to be used against an intruder.

Hearings Before the Subcomm. on Crime of the Comm. on the Judiciary, 94th Cong. 1774 (1975) (statement of Dee Helfgott, Coordinator, Coalition for Handgun Control of Southern California, Inc.).

[22]. Zimring & Hawkins, supra note 1, at 190.

[23]. See, e.g., Citizen's Guide, supra note 3; Norval Morris & Gordon Hawkins, An Honest Politician's Guide to Gun Control (1969); George Newton & Franklin Zimring, Firearms and Violence in American Life (1969); Franklin Zimring, Fun and Games with Guns and Statistics, 1968 Wis. L. Rev. 1113.

[24]. See Zimring & Hawkins, supra note 1, at 199.

[25]. See id. at 201.

[26]. Colin Greenwood & Joseph Magaddino, Comparative Cross-Cultural Statistics, in Restricting Handguns 39 (Don Kates ed., 1979). At the time this was written, Greenwood held the rank of Chief Inspector of the West Yorkshire Constabulary. He retired as a Chief Superintendent. See also Colin Greenwood, Firearms Control: A Study of Armed Crime and Firearms Control in England and Wales (1972) [hereinafter Firearms Control].

[27]. See generally Gary Kleck & Marc Gertz, Armed Resistance to Crime: The Nature and Prevalence of Self-Defense with a Gun, 86 J. Crim. L. & Criminology 150 (1995); Philip Cook et al., The Gun Debate's New Mythical Number: How Many Defensive Uses Per Year? 16 J. Pol'y Analysis 463 (1997).

[28]. After John Lennon was murdered, a widely circulated hippie bumper sticker read: "Imagine the world without guns." What was meant, we suppose, was: "Imagine the world without coercion"--not the same thing. Of course one need not "imagine" the world without guns, one need only remember it. Practical, portable firearms were not available in Europe until the fifteenth century, see Archer Jones, The Art of War in the Western World 151-53 (1987), and much later in East Asia. Notwithstanding, there was homicidal violence in kind and quantity that still seems impressive in modern times. See generally Roger Lane, Murder in America 14 (1997); Ted Robert Gurr, Historical Trends in Violent Crime: Europe and the United States, in 1 Violence in America 21-54 (Ted Robert Gurr ed., 1989) [hereinafter Historical Trends]; Ted Robert Gurr, Historical Trends in Violent Crime: A Critical Review of the Evidence, in 3 Crime and Justice: An Annual Review of Research 295 (1981) [hereinafter Critical Review].

[29]. Zimring & Hawkins, supra note 1, at 21.

[30]. Not, apparently, "the United Kingdom," which consists also of Northern Ireland and Scotland.

[31]. See Zimring & Hawkins, supra note 1, at 27-28.

[32]. See id. at 35-35.

[33]. Id. at 37.

[34]. See id. at 38.

[35]. See id. at 38-39.

[36]. Id. at 50.

[37]. Only 10% of rapes involve the perpetrators using a firearm. See Patsy A. Klaus & Marshall Deberry, The Crime of Rape, Bureau of Justice Statistics Bull., Mar. 1985, at 4 tbl.11.

[38]. Based on Interpol Statistics for the year 1984, the American rape rate was 13.2 times greater than that of England and Wales. See Carol A. Kalish, Bureau of Justice Statistics, Special Rep.: International Crime Rates 3 tbl.4 (May 1988). Based on 1988 survey evidence, the U.S. victimization rate for sexual assault was 23 times greater than the England/Wales rate. See Jan J.M. van Dijk, On the Uses of Local, National and International Crime Surveys, in Australian Institute of Criminology, International Trends in Crime: East Meets West 137, 147 fig.6 (1992). Despite the practice of some modern unisex rape statutes, it is not clear that "sexual assault" always or even usually equals "rape"; either or both may be intended to include forcible sodomy as well. But "sexual assault" evidently does mean serious predation of some kind since van Dijk contradistinguishes it from mere "sexual incidents," also covered by the survey. See id.

[39]. See John Kaplan et al., Criminal Law 29 (1996).

[40]. See id.

[41]. See id.

[42]. See Bureau of Justice Statistics, Sourcebook of Criminal Justice Statistics 319 tbl.3.11 (1996); see also United Nations, 1995 Demographic Yearbook (1995).

[43]. See Universal Almanac 294 (1997).

[44]. See American Almanac 1994-95 tbl.306 & tbl.307 (1995) (reporting FBI data).

[45]. See James Q. Wilson & Richard J. Herrnstein, Crime and Human Nature 461 (1985).

[46]. Lane, supra note 28, at 327.

[47]. Elijah Anderson, The Code of the Streets, Atlantic Monthly, May 1994, at 80.

[48]. See generally David Courtwright, Violent Land: Single Men and Social Disorder from the Frontier to the Inner City (1996).

[49]. See Daniel D. Polsby, Ending the War on Drugs and Children, 31 Val. U. L. Rev. 537, 543 (1997).

[50]. See John DiIulio, Jr., The Question of Black Crime, 117 Pub. Interest 3 (1994).

[51]. C. Vann Woodward, The Age of Reinterpretation, Am. Hist. Rev., Oct. 1960, reprinted in The Future of the Past 93 (1989).

[52]. Oft-repeated dictum of University of Chicago law faculty, see e.g., Frank H. Easterbrook & Daniel R. Fischel, Contract and Fiduciary Duty, 36 J.L. & Econ. 425, 434 (1993); Richard A. Epstein, Common Law, Labor Law, and Reality: A Rejoinder to Professors Getman and Kohler, 92 Yale L.J. 1435, 1435 (1983).

[53]. Oft-repeated but unpublished precept of Professor Stephen Morse of the University of Pennsylvania.

[54]. The homicide data were collected by Center for Disease Control employees Etienne G. Krug, Kenneth E. Powell and Linda Dahlberg in an unpublished manuscript, Firearm-Related Death in the United States and 35 Other High- and Upper-Middle-Income Countries. They were shared with Professor Kleck under the stipulation that he not publish them. The bare conclusions from his analysis are set out in Kleck, supra note 14, at 254-55.

[55]. See Kleck, supra note 14, at 359.

[56]. See Joyce Lee Malcolm, The Link Between Firearms and Violence: The English Example, Address at the American Society of Criminology 1997 Annual Meeting (Nov. 22, 1997).

[57]. Studies showing this for various time periods in the three countries are: Metro Chicago Information Center, Summary of Neighborhood Crime Perceptions from the 1991-1995 Metro Survey (1995) [hereinafter Metro Chicago Survey]; James Wright et al., Under the Gun: Weapons, Crime and Violence in the United States 107 (1983); Gary Kleck, The Relationship Between Gun Ownership Levels and Rates of Violence in the United States, in Firearms and Violence (Don Kates ed., 1986); Gary Kleck & Britt E. Patterson, The Impact of Gun Control and Gun Ownership Levels on City Violence Rates, 9 J. Quant. Criminology 249-87 (1993); Richard Munday, The Right to Arms [in England], Salisbury Rev., Summer 1997, at 6, 8; Philip C. Stenning, Gun Control--A Critique of Current Policy, 15 Pol'y Options 13, 14 (1994); Hans Toch & Alan J. Lizotte, Research & Policy: The Case of Gun Control, in Psychology and Social Policy (Peter Suedfeld & Philip E. Tetlock eds., 1990).

[58]. Toch & Lizotte, supra note 57, at 232. As Ramsey Clark put it: "A gun emboldens you. Suppose your victim is bigger than you. Guns more than equalize. With a gun you are eight feet tall." Clark, supra note 21.

[59]. Toch & Lizotte, supra note 57, at 234.

[60]. Id. at 234 n.10.

[61]. One might say, of course, that it is an apples-to-oranges affair to compare the State of Idaho with the City of Detroit (population of each: 1.1 million) or the State of Maine (population 1.2 million) with the City of Dallas (population 1 million), or the State of Wyoming (population 480,000) to the City of New Orleans (population 496,000). See American Almanac, supra note 44, at 27 tbl.26 & 44-45 tbl.46. After all, rural states and teeming cities are very different places. What can one learn from comparing them? A fair point, but it needs extending. Comparative arguments of this kind are liable to cover over very important assumptions that need to be validated before one can learn anything useful from the analysis. New Orleans and Wyoming are indeed very different places, as are the State of Idaho and the City of Detroit. And of course the same is true of New York City and London, and Sydney and Los Angeles. For some purposes, at least, there are surely greater affinities between Americans living in big cities and their rural compatriots than there are between American city folk and foreigners. Maybe these issues of cultural cohesion go no further than whether Johnny Carson is funny, Vegemite is edible, or whether soccer is more fun to watch than actual football; or maybe there are larger affinities that could make crime studies based on comparison suggestive.

[62]. Residents of suburban low-crime communities are much more likely to own handguns than those living in the city. See Metro Chicago Survey, supra note 57, at 6; see also Robert A.I. Munday & Jan A. Stevenson, Guns & Violence: The Debate Before Lord Cullen 279 (1996).

[63]. Firearms stock figures in the table are taken from Kleck, supra note 54. For 1995 figures, see infra note 64. Murder rates in the table are taken from Bureau of Justice Statistics, U.S. Dep't of Justice, Sourcebook of Criminal Justice Statistics--1989, 365 tbl.3.118 (Timothy J. Flanagan et al. eds., 1989) [hereinafter Bureau of Justice Statistics 1989] and Bruce Taylor, Changes in Criminal Victimization 1994-1995, Bureau of Just. Statistics Nat'l Crime Victimization Survey, Apr. 1997, at 6 tbl.A. For the sake of simplicity, we are adopting the convention of "no depreciation" in the firearms stock, which is certainly false but probably not seriously misleading. Some number of operable weapons, no one knows how many, do become inoperable in the course of any given year. We think it is unlikely that the number is large, and may be pushed toward insignificance because of unlawful additions to the civilian gun stock, smuggled and illegally manufactured weapons, that go unenumerated by government. It goes without saying that, by assuming a high enough depreciation rate, one could make an apparently increasing national inventory of firearms into an actually decreasing one. We are aware of no friend of the instrumentality theory who has chosen to join issue on this risky basis.

[64]. References to 1995 gun ownership are extrapolations from Kleck, supra note 14.

[65]. See id.; Bureau of Justice Statistics 1989, supra note 63; Taylor, supra note 63.

[66]. See Kleck, supra note 14; see also Table 1: Murder Rates in Relation to Firearms Inventory, 1975-1996, supra page 984.

[67]. See Bureau of Justice Statistics 1989, supra note 63; Taylor, supra note 63; see also Table 1: Murder Rates in Relation to Firearms Inventory, 1975-1996, supra page 984.

[68]. See supra note 64.

[69]. See Taylor, supra note 63.

[70]. See also Chicago Police Department, Detective Division, Murder Analysis, 1965-1996 (annual reports). Domestic homicide is defined as homicide between spouses, "common law" spouses, siblings, parents and children, grandparents and grandchildren, uncle or aunt and nephew or niece, cousins, in- laws, step relatives, foster relatives, boyfriend and girlfriend, ex-boyfriend and girlfriend, child or mother's boyfriend, and friends.

[71]. In 1977 and 1978 the homicide rate had dropped to 8.8 per hundred thousand population; by 1980, it was up to 10.2; by 1983 and 1984, it was down to 7.9; by 1991, it peaked at 9.8; by 1995, it was back down to 8.2. According to the latest available FBI estimate, in 1996 it decreased another 11% from 1995. See Uniform Crime Reports: 1996 Preliminary Annual Release (last modified June 1, 1997) http://www3.phillynews.com/data/ucr96/ucr96.html

This would indicate a figure in the neighborhood of 7.3, which would make it the lowest national murder rate recorded since 1968.

[72]. Alfred Blumstein, Youth Violence, Guns and the Illicit-Drug Industry, 86 J. Crim. L. & Criminology 10, 11 (1995).

[73]. Clark, supra note 21, at 108.

[74]. See Lane, supra note 28, at 326; see also Chicago Police Department, supra note 70; Figure 1: Total Homicides and Domestic Homicides, City of Chicago, 1965-1996, supra page 985.

[75]. See Philip J. Cook & John H. Laub, The Unprecedented Epidemic in Youth Homicide, in Crime and Justice: An Annual Review of Research 13 tbl.6.1 (Mark Moore & Michael Tonry eds., forthcoming 1998) (manuscript on file with authors).

[76]. See Bureau of Justice Statistics, U. S. Dep't of Justice, Sourcebook of Criminal Justice Statistics--1993 at 388 tbl.3.135 & fig.3.7 (1993).

[77]. See Center for Disease Control, Homicides Among 15-19 Year Old Males--United States 1963-1991, 43 Morbidity and Mortality Wkly Rep. 725, 726-27 (1994) (noting drastically rising homicide among inner-city youth and suggesting that "the immediate and specific causes" may include "the recruitment of juveniles into drug markets"); see also Blumstein, supra note 72, at 11-12; David M. Kennedy et al., Youth Violence in Boston: Gun Markets, Serious Youth Offenders, and a Use-Reduction Strategy, Law & Contemp. Probs., Winter 1996, at 147, 159-60; Polsby, supra note 49; Franklin Zimring & Gordon Hawkins, Concealed Handguns: The Counterfeit Deterrent, Responsive Community, Spring 1997, at 46, 52.

[78]. See John R. Lott, Jr. & David B. Mustard, Crime, Deterrence, and Right-to-Carry Concealed Handguns, 26 J. Legal Stud. 1 (1997).

[79]. See, e.g., sources cited infra notes 81-84.

[80]. See Bob Greene, Wild West Mentality Spreads from Streets to Legislatures, Chi. Trib., Oct. 4, 1995, at Tempo 1.

[81]. See Don McLeese, Handgun Law Draws Ambivalence; Handgun Law Draws Ambivalent Reactions, Austin Am.-Statesman, Dec. 7, 1995, at B1.

[82]. See Editorial, Turning Florida into the Wild West, Daytona Beach Evening News, Apr. 10, 1985, at 4A (concealed carry is a "dumb idea" that would "put more and more innocent, law abiding citizens in daily danger of being shot--or shooting themselves or loved ones accidentally"); Laurie Hollman, Measures to Loosen Controls on Guns Pass in House Panel, St. Petersburg Times, Apr. 16, 1985, at 6B.

[83]. See Guy Friddell, Preparing for the Perils of Policing Dodge City, Virginian-Pilot, July 13, 1995, at B1.

[84]. See Paul Clark, Boosting City Hall Security, Asheville Citizen- Times, July 27, 1995, at 1B ("This is going to be Dodge City. We're going to end up calling ourselves Bubbaville...."); Leni Sitnick, Law to Permit Concealed Guns Makes Bad Public Policy, Asheville Citizen-Times, July 23, 1995, at 7B (arguing that Small Town, U.S.A., will be turned into Dodge City).

[85]. Shades of the nuclear disarmament problem: this is an example of what is called a "first mover" situation, in which the player who moves first obtains a significant strategic advantage over his opponent or opponents. See Daniel D. Polsby, Equal Protection, Reason, Oct. 1993, at34. See generally Paul Stephan, Barbarians Inside the Gate: Public Choice Theory and International Economic Law, 10 Am. U. J. Int'l L. & Pol'y 745, 758 (1995).

[86]. See, e.g., sources cited supra notes 81-84.

[87]. See Clayton E. Cramer & David B. Kopel, "Shall Issue": The New Wave of Concealed Handgun Permit Laws, 62 Tenn. L. Rev. 679 (1995), for an early review of these laws and the experience under them. See Lott & Mustard, supra note 78, for a later review concentrating on the empirical effects of the laws.

[88]. Texas: As of late 1996 Texas had issued 111,408 licenses. See Richard Connelly, Handgun Law's First Year Belies Fears of 'Blood in the Streets', Tex. Law., Dec. 9, 1996, at 2. As of early 1997 another roughly 5,000 concealed carry licenses had issued, swelling the number to "more than 116,000." Editorial, Shootout in Mild West, Austin Am.-Statesman, Feb. 2, 1997, at J2.

Florida: As of October 31, 1997, there were 210,977 concealed carry licenses outstanding. See Concealed Weapons/ Firearms Monthly Report (visited October 31, 1997) <http://www.dos.state.fl.us/dol/stats/cw_monthly.html> (This site is updated periodically.).

Connecticut: According to a Connecticut State Police computer list, 115,877 CCW licensees were outstanding as of August 1991. Interview with Greg Miller and Don Bernardo (Aug. 8, 1994) (Their information was based on information released on a Freedom of Information Act request obtained by order of Superior Court, Judicial District of Hartford, New Britain at Hartford, CV- 91-0702142S.).

North Carolina: See Thom Goolsby, Concealed Weapons Advocates Were Right: Crime Didn't Go Up [in N.C.], Chapel Hill Herald (N.C.), May 6, 1997, at 4 (reporting that there are 20,082 licensees in North Carolina).

Idaho: 4,000 Ada county residents (Boise area) have CCWs; around 19,000 statewide. See Phil Sahm, Uniform Standard Covers Idaho's Concealed Weapons Permit Process, Idaho Statesman, Jan. 20, 1995, at 14A.

Iowa: "20,587 people [are] licensed to walk around in Iowa with a concealed weapon--and that doesn't include police or soldiers or guards or private detectives." Michael Gardiner, Handguns Not Just a Big-City Problem, USA Today, Jan. 13, 1994, at 9A. Iowa's 1992 population was 2,812,448. See Wright et al., supra note 57, at 145. That computes to one licensee for every 137 Iowans.

Washington: The State Firearms License Bureau reported that as of October 1991 (their latest available figure) a total of 198,163 licenses to carry handguns concealed were outstanding. Based on Washington's 1990 population of 4,866,692 this is one license for every 25.7 residents. "As of 1993 there were 241,606 licenses outstanding in Washington." Cramer & Kopel, supra note 87, at 689.

Indiana: The Indiana State Police report that they currently had outstanding 220,623 permits to carry handguns concealed. Based on Indiana's 1990 population of 5,544,159, this is one license for every 25 residents.

Georgia: There is no state-wide data base. The Fulton County (Atlanta) court clerk in charge of issuance in that county estimates the number of outstanding permits to carry handguns concealed at 190,800 state-wide. Based on Georgia's 1990 population of 6,478,216, this is one license for every 34 residents.

New Hampshire: There is no state-wide database. State police estimate that the number state-wide is 80,000. Based on New Hampshire's 1990 population of 1,109,252 this is one license for every 13.9 residents.

Pennsylvania: 362,142 CCW licenses were outstanding as of January 1992 (about three percent of the population). Not all were issued to residents.

[89]. See Connelly, supra note 88 (analyzing results of Texas law); Terry Flynn, Gun-Toting Kentuckians Hold Their Fire, Cincinnati Enquirer, June 16, 1997, at A1 (analyzing results of new Kentucky law); Goolsby, supra note 88 (analyzing results of North Carolina law); Police Say Concealed Weapons Law Has Not Brought Rise in Violence, Palm Beach Post, July 26, 1988, at 1A; Records Say Licensed Gun Owners Are Least of Florida's Crime Problem, Tallahassee Democrat, Nov. 4, 1990, at 6B.

[90]. Michael Warren, Concealed Weapons Owners No Trouble, Gainesville Sun, Nov. 4, 1990, at 1A (quoting Florida Representative Ron Silver); see also Steve Patterson, Concealed Weapons Law Opponents Still Searching for Ammunition, Fla. Times-Union, May 9, 1988, at A-1 (president of Florida Chiefs of Police who opposed law admits that, despite his organization's attempts to document problems from every department in the state, they have found none); Police Say Concealed Weapons Law Has Not Brought Rise in Violence, Palm Beach Post, July 26, 1988, at 1A (executive director of Florida Chiefs of Police admits the same and so does official of the Florida Sheriff's Association).

[91]. Connelly, supra note 88.

[92]. Flynn, supra note 89.

[93]. Thom Goolsby, Stop Shooting Down Concealed-Weapon Law, Greensboro News & Rec. (N.C.), May 11, 1997, at F3 (quoting Fayetteville police chief Ron Hansen).

[94]. See Dan Black & Daniel Nagin, Do "Right-to-Carry" Laws Deter Violent Crime? 27 J. Legal Stud. 209 (1998); Hashem Dezhbakhsh & Paul Rubin, Lives Saved or Lives Lost: The Effects of Concealed Handgun Laws on Crime (Jan. 1998) (unpublished manuscript, presented at American Economic Association meetings in Chicago); Jens Ludwig, Do Permissive Concealed Carry Laws Deter Violent Crime (1997) (unpublished manuscript, on file with author).

[95]. See Ludwig, supra note 94. In principle, it should be possible to measure the effect of changes in concealed carry laws on actual (and perceived) gun-carrying behavior, but this research has not yet been carried out.

[96]. Violence Policy Center, supra note 17; see also Josh Sugarmann & Kristen Rand, Cease Fire: A Comprehensive Strategy to Reduce Firearms Violence 4 (1994) ("The majority of firearm murder stems from arguments that turn deadly because of ready access to a firearm."); Josh Sugarmann, The NRA Is Right--But We Still Need to Ban Handguns, Wash. Monthly, June 1987, at 11 ("[M]ost [handgun murders] are not caused by predatory strangers. Handgun violence is usually the result of people being angry, drunk, careless, or depressed--who just happen to have a handgun around.").

[97]. Zimring & Hawkins, supra note 1, at 61.

[98]. Id. at 16.

[99]. Id. at 61.

[100]. Id. at 78.

[101]. Deane Calhoun, From Controversy to Prevention: Building Effective Firearm Policies, Inj. Prevention Network Newsl., Winter 1989-90, at 15.

[102]. Lindsay, supra note 19; see also National Coalition, supra note 18 (citing Lindsay and asserting that "each year" thousands of "gun murders [are] done by law-abiding citizens who might have stayed law-abiding if they had not possessed firearms" (emphasis in original) and "that most murders are committed by previously law abiding citizens where the killer and the victim are related or acquainted"); Edwards, supra note 21.

[103]. See Calhoun, supra note 101 (Most murderers "are neither felons nor crazy," but rather "people involved in family fights and fights over jobs and money, and people who are sad or depressed."); Katherine Kaufer Christoffel, Toward Reducing Pediatric Injuries from Firearms: Charting a Legislative and Regulatory Course, 88 Pediatrics 294, 300 (1991) ("[M]ost shootings are not committed by felons or mentally ill people, but are acts of passion that are committed using a handgun that is owned for home protection."); Daniel W. Webster et al., Reducing Firearms Injuries, Issues In Sci. & Tech., Spring 1991, at 73 ("[M]ost [murderers] would be considered law-abiding citizens prior to their pulling the trigger.").

[104]. See Sheilah Hodgins, Mental Disorder, Intellectual Deficiency, and Crime, 49 Archives of Gen. Psychiatry 476 (1992), for a collection and analysis of particular studies.

[105]. Pekka Santilla & Jaana Haapasalo, Neurological and Psychological Risk Factors Among Young Homicidal, Violent, and Nonviolent Offenders in Finland, 1 Homicide Stud. 234 (1997). The authors continue:

For example, Lewis et al. (1985) found that in a small sample of nine homicidal offenders, neurological impairment was present in several subjects. In a study comparing juvenile murderers with other violent delinquents and nonviolent delinquents, Lewis et al. (1988) showed that the murderers did not differ from the other violent delinquents on neuropsychiatric variables but that they were more impaired neuropsychiatrically than the nonviolent offenders (also, see Nestor, 1992). Support for neurological abnormalities in association with violent behavior has been found in extensive reviews of the literature.

Id. (some citations omitted).

[106]. Id. at 237 (citations omitted).

[107]. Id. at 247 (citations omitted). Though we have generally deleted the citations, the quoted language included references to ten different studies or literature reviews. In the Finnish study, however, differences between violent offenders and murderers were found to be somewhat less marked than in some of the earlier studies.

[108]. Ronald M. Holmes & Stephen T. Holmes, Murder in America 8-9 (1994).

[109]. See, e.g., D. Mulvihill et al., Crimes of Violence: Report of the Task Force on Individual Acts of Violence 532 (1969) (data run of murder arrestees nationally over a four-year period in the mid 1960s found 74.7% to have had prior arrest(s) for violent felony or burglary); Federal bureau of Investigation, U. S. Dep't of Justice, Uniform Crime Report 38 (1971) (in data run of murder arrestees over a one-year period 77.9% had priors). In two Bureau of Justice Statistics special reports covering late 1980s national samples, 76.7% of murder arrestees had criminal histories as did 78% of defendants in murder prosecutions. See Bureau of Justice Statistics, U. S. Dep't of Justice, Special Rep.: Murder in Families 5tbl.7 (1994); Bureau of Justice statistics, U. S. Dep't of Justice, Special Rep.: Murder in Large Urban Counties, 1988 (1993); see also R. Narloch, Cal. Bureau of Criminal Statistics, Criminal Homicide in California 53-54 (1973); A. Swersey & E. Enloe, Homicide in Harlem 17 (1975) ("We estimate that the great majority of both perpetrators and victims of assaults and murders had previous arrests, probably over 80% or more.").

[110]. "After a long plateau at around 10 percent, the juvenile share of homicide arrests started climbing in the mid-1980s, [peaking at] 17 percent by 1994." Cook & Laub, supra note 75.

[111]. See Federal Bureau of Investigation, U.S. Dep't of Justice, Uniform Crime Report 42 (1975).

[112]. See Mark Cooney, The Decline of Elite Homicide, 35 Criminology 381 (1997).

[113]. See Bureau of Justice Statistics, U. S. Dep't of Justice, Special Rep.: Probation and Parole Violators in State Prison (Aug. 1995) (surveying state prison inmates in 1991).

[114]. See DiIulio, supra note 50, at 16.

[115]. "[M]ost violent acts escape the attention of authorities and are thus not made a part of official written records...." Gary Kleck, Policy Lessons from Recent Gun Control Research, L. & Contemp. Probs., Winter 1986, at 35, 40.

[116]. See T. Markus Funk & Daniel D. Polsby, Some Distributional Consequences of Expunging Juvenile Delinquency Records: The Problem of Lemons, 52 Wash. U. J. Urb. & Contemp. L. 161 (1997).

[117]. See Kennedy et al., supra note 77, at 159-60.

[118]. Webster et al., supra note 103.

[119]. Murray A. Straus, Domestic Violence and Homicide Antecedents, 62 Bull. N.Y. Acad. Med. 446, 454 (1986).

[120]. See id. at 457.

[121]. See Kleck, supra note 115, at 41.

[122]. See David M. Kennedy, Pulling Levers: Chronic Offenders, High- Crime Settings, and a Theory of Prevention, 31 Val. U. L. Rev. 455 (1997) ("[D] omestic violence offenders, at least those who come to the attention of the criminal justice system, tend to have robust [prior] offending histories.").

[123]. Straus, supra note 119, at 450.

[124]. See, e.g., Lane, supra note 28, at 324, 341-53.

[125]. See Kennedy et al., supra note 117, at 159.

[126]. Id. at 158.

[127]. Id. at 162.

[128]. See Cook & Laub, supra note 75, at 13.

[129]. H. Range Hutson et al., Adolescents and Children Injured or Killed in Drive-By Shootings in Los Angeles, 330 New Eng. J. Med. 324, 325 (1994).

[130]. See Carl A. Soderstrom et al., Psychoactive Substance Use Disorders Among Seriously Injured Trauma Center Patients, 277 JAMA 1769, 1771 (1997). Other trauma care center studies disgustedly characterize urban knife and bullet wounds as "a chronic recurrent disease peculiar to unemployed, uninsured law breakers." R. Stephen Smith et al., Recidivism in an Urban Trauma Center, 127 Archives of Surgery 668, 670 (1992) (describing the conclusions in D. W. Sims et al., Urban Trauma: A Chronic, Recurrent Disease, 29 J. Trauma 940 (1989); M.C. Morrisey et al., The Incidence of Recurrent Penetrating Wound Trauma in an Urban Trauma Center, 31 J. Trauma 1536 (1991)).

[131]. Swersey & Enloe, supra note 109.

[132]. See Chicago Police Department, supra note 74, at 21.

[133]. See Bureau of the Census, U.S. Dep't of Com., Statistical Abstract of the United States: National Data Book 1995, at 202 tbl.314 (1995).

[134]. See, e.g., Frank Zimring, Is Gun Control Likely to Reduce Violent Killings?, 35 U. Chi. L. Rev. 721 (1968).

[135]. Cf. Lambert v. California, 355 U.S. 225 (1957) (holding that the defendant's ignorance of a felon registration law did excuse failure to comply because the "criminal act" was "wholly passive").

[136]. See Tom Tyler, Why People Obey the Law 60 (1990).

[137]. Many more people in Chicago signed up for their auto registration stickers once it became possible to do so over the telephone rather than waiting in a long line at City Hall.

[138]. See Tyler, supra note 136.

[139]. See John Sloan et al., Firearms Regulation and Rates of Suicide, 322 New Eng. J. of Med. 369, 369-73 (1990); John Sloan et al., Handgun Regulations, Crime, Assaults, and Homicide: A Tale of Two Cities, 319 New Eng. J. of Med. 1256, 1256-62 (1988).

[140]. See Zimring & Hawkins, supra note 1, at 44-50.

[141]. See, e.g., Susan Baker, Without Guns Do People Kill People? 75 Am. J. Pub. Health 587 (1985); Charles M. Callahan & Frederick P. Rivara, Urban High School Youth and Handguns: A School-Based Survey, 267 JAMA 3038, 3042 (1992); Judith Cohen Dolins & Katherine Kaufer Christoffel, Reducing Violent Injuries: Priorities for Pediatrician Advocacy, 94 Pediatrics 638, 648 (1994); Roberta K. Lee et al. Incidence Rates of Firearm Injuries in Galveston, Texas, 1979-1981, 134 Am. J. Epidemiology 511, 520 (1991); James A. Mercy et al. Public Health Policy for Preventing Violence, 12 Health Aff. 7, 17-19 (1993); James A. Mercy et al. Firearm Injuries: A Call for Science, 319 New Eng. J. Med. 1283, 1283-84 (1988); Stephen P. Teret & Garen J. Wintemute, Policies to Prevent Firearms Injuries, 12 Health Aff. 96, 101-06 (1993); Jon S. Vernick & Stephen P. Teret, Firearms and Health: The Right to Be Armed with Accurate Information About the Second Amendment, 83 Am. J. Pub. Health 1773, 1777 (1993); Webster et al., supra note 103, 73-75; Daniel W. Webster & Modena E. H. Wilson, Gun Violence Among Youth and the Pediatrician's Role in Primary Prevention, 94 Pediatrics 617, 622 (1994); Garen J. Wintemute, Closing the Gap Between Research and Policy: Firearms, Inj. Prevention Network Newsl., Winter, 1989-90.

[142]. Much of continental Europe (but not Switzerland or England) had severely restricted firearms ownership during the long period of autocracy prior to the French Revolution. See Lee Kennett & James LaVerne Anderson, The Gun in America: The Origins of a National Dilemma, 3-34 (1975); Don B. Kates, Jr., Handgun Prohibition and the Original Meaning of the Second Amendment, 82 Mich. L. Rev. 203, 233 (1983) (France). By the late-nineteenth century, however, autocracy had largely been replaced with at least semi-popular government in western Europe, with French-style universal military service, "and much of the restrictive arms legislation was rescinded." Kennett & Anderson, supra, at 32 (citing particularly a French-language source for France). See Firearms Control, supra note 26, at ch. 1, for a review of late- nineteenth century European law.

[143]. See generally Firearms Control, supra note 26, chs. 1-3; Critical Review, supra note 28; Historical Trends, supra note 28; Eric Monkkonen, Diverging Homicide Rates: England and the United States, in Violence in America, supra note 28; Frank Morn, Firearms Use and Police: An Historic Evolution of American Values, in Firearms and Violence, supra note 57, at 496- 501.

[144]. Ian Hogg, German Pistols and Revolvers, 1871-1945 (Harrisburg, Stackpole, 1971) (listing, inter alia, Becker & Hollander of Suhl and Stendawerke G.m.b.H, at 10; Aktiengesellschaft Lignose of Berlin and Simplex, at 14; Gustav Bittner, at 15; Ludwig Lowe or Berlin, at 17; DWM (which is still in operation), at 18; Buco DRGM, at 19; V.C. Schilling, C.G. Haenel, and J.P. Sauer, all of Suhl, at 20, of which the Sauer firm is the only survivor; Rheinische Waffen-und-Munitionsfabrik of Cologne, at 22; Rheinische Metallwaren-und Maschinenfabrik, at 23; Jager-Pistole DRP Angem, at 34; Kommer Waffenfabrik Zella Mehlis, at 35; Fritz Langenhan of Suhl, at 37; Theodore Bergmann Gaggenau Waffenfabrik Suhl which later became Aktiengesellschaft Lignose of Berlin, at 41; Fritz Mann and Mann-Werke A.G. of Suhl, at 44-45; Deutsche Werke Aktiengesellschaft Werk Erfurt, at 63; Waffenfabrik Bern and Heinrich Krieghoff Waffenfabrik, at 67; Friedrich Pickert Waffenfwork of Zella Mehlis, at 78; Romerwerke of Suhl, at 92; Schwarzlose GmbH Berlin, at 104; Waffenfabrik Simson & Co., at 108; Franz Stock, Berlin, at 110).

[145]. Raymond Caranta, Pre-WWII Pocket Pistols, in Gun Digest--1997, at 140 (Ken Warner ed.).

[146]. See Editorial, The Menace of the Pistol, Boston Advertiser, 1910, reprinted in 2 American Institute of Criminal Law 23, 93 (1911). For the nineteenth- and twentieth-century development of "Saturday Night Special" and other American handgun bans or licensing schemes, see Don B. Kates, Toward a History of Handgun Prohibition in the United States, in Restricting Handguns 7 (Don B. Kates ed., 1979).

[147]. Monkkonen, supra note 143, at 81-82.

[148]. See Lane, supra note 28.

[149]. John Wood Osborne, The Silent Revolution 9 (1970); cf. David Wooton, Disarming the English, London Rev. Books, July 21, 1994, at 20 ("I cannot now stir a mile from my own house after sunset without one or two servants with blunderbusses," wrote Horace Walpole in 1782, "a footpad or a highwayman might be around any turn in the road.").

[150]. 10 & 11 Geo. 5, ch. 43 (current version within Firearms Act, 1968, ch. 27).

[151]. See Kennett & Anderson, supra note 142, at 33; Malcolm, supra note 56; see also Don B. Kates, Guns, Murders and the Constitution: A Realistic Assessment of Gun Control 39 (1990) [hereinafter Guns, Murders and the Constitution]:

The different purposes of [post World War I] European versus American laws are evidenced by their diametrically opposite patterns: many of the "Saturday Night Special" laws American states enacted to deal with 19th Century crime banned all but standard military-issue revolvers (i.e. the very expensive large, heavy Colt). In stark contrast, such military caliber arms were the first guns banned in post-WWI Europe, the purpose being to disarm restive former soldiers and the para-military groups they formed.

Id.

[152]. See Daniel Polsby & Don Kates, Of Holocausts and Gun Control, 75 Wash. U. L. Q. 1237, 1240 (1998).

[153]. See David B. Kopel, The Samurai, the Mountie and the Cowboy (1992); Critical Review, supra note 28; Historical Trends, supra note 28.

[154]. Firearms Control, supra note 26, at 243.

[155]. See Guns, Murders and the Constitution, supra note 151, at 40; Kopel, supra note 153, at 91-93. "[I]n 1974 the American rate was 40 times the English; 15 years ...later, the American rate was only ten times greater." Guns, Murders and the Constitution, supra note 151, at 40. Though this convergence reflects some decrease in American homicide, it is primarily due to a steady increase in English homicide.

[156]. See discussion and sources cited in Mayer Freed & Daniel Polsby, Just Cause for Termination Rules and Economic Efficiency, 38 Emory L. J. 1097, 1140 (1989).

[157]. See Kleck, supra note 9, at 223-68; Lawrence Southwick, Do Guns Cause Crime? Does Crime Cause Guns? A Granger Test, 25 Atlantic Econ. J. 256, 262-65 (1997).

[158]. Wright et al., supra note 57, at 138.