Michigan Law Review
Book Review, 83 (1985): 954.
Posted for Educational use only. The printed edition remains canonical. For citational use please visit the local law library or obtain
a back issue.
TWO NEW BOOKS ON GUNS
UNDER THE GUN: WEAPONS, CRIME AND VIOLENCE IN AMERICA
By James D. Wright, Peter H. Rossi and Kathleen Daly, with the assistance of Eleanor Weber-Burdin.
New York: Aldine Publishing Co. 1983. Pp. xviii, 342. $24.95.FIREARMS AND VIOLENCE: ISSUES OF PUBLIC POLIC
Edited by Don B. Kates, Jr. Cambridge, Mass.: Ballinger Publishing Co. 1984. Pp. xxxiii, 571. $38.Franklin E. Zimring *
Copyright 1985 by the Michigan, Law Review Association & Franklin E. Zimring
A striking fact about the statistical and analytic literature on firearms and violence in the United States is that there isn't much. There is very little research and scholarly writing of any quality on the topic, and only a small fraction of that would meet the most primitive standards of quality control. While specialists in any field frequently complain that their per topic has received insufficient attention from social and policy sciences, the dearth of responsible discussions of guns and gun control is both easy to document and worthy, itself, of sustained study.
The status of firearms and violence as an unwelcome stepchild in contemporary policy research grows out of two social phenomena. First gun control is an ideologically dominated and emotionally charged issue, one of many in this country where both sides of a heated policy debate feel no great need for factual support to buttress foregone conclusions. Second, in the case of guns, the disregard for factual evidence by partisans is not counterbalanced by a tradition of investigation by noncombatants. The study of firearms and violence has no natural constituency among the academic disciplines. There are no departments of guns in our universities, and we lack research centers or institutes specializing in such matters attached to schools of public health, programs of public policy, or departments of sociology.
The lack of an academic connection compounds problems generated by the pervasive ideological cast of the gun control debate. Even behavioral research on pornography, for example, is more common than on guns. Alcohol studies, while not securely under a single academic department, have been institutionalized in the research community [Page 955] as an interdisciplinary topical specialty. In the United States, there are certified experts within the academy in highway safety, drug abuse, and wife beating. There are no experts on guns.
As bad as things are in the scientific study of firearms and violence, conditions were worse until quite recently. Twenty years ago, during a period of sustained interest in federal and state gun control legislation, there was literally no current scholarship on the relationship between guns and the incidence or consequences of interpersonal violence and no work was in progress. Since the report of the National Commission on the Causes and Prevention of Violence in 1969, [1] a small number of scholars have begun to pay sustained attention to the social and policy issues raised by firearms and governmental attempts at their control. While no one I know specializes in these issues exclusively, it is now possible at least to fill a small seminar room with serious students of the topic. Ten years ago, this could not have been done.
The two books under review are indications of how far the study of firearms and violence has come in recent years, as well as how far it has yet to go. Under the Gun is the collaborative effort of three establishment social scientists associated with a large American university and supported by a research grant from the federal government. Firearms and Violence is a more counter-culture document, a 600-page collection of essays and studies commissioned and edited by a San Francisco attorney, Don B. Kates, who is fast acquiring a reputation as one of the more articulate, bright, and individualistic opponents of public law regulation of firearms.
While divergent in pedigree, these two efforts have much in common. Both books are skeptical of the potential effectiveness of firearms control laws in reducing the death toll from violent crime. Both books, in contrast to previous mainstream social science, dispute that the link between firearm (particularly handgun) availability and the death rate from violence has been demonstrated. Both books emphasize statistical materials and quantitative methods, unlike most previous scholarly writing on firearms regulation and almost all polemics on the topic. Both books review a substantial proportion of the available literature on the relationship between firearms and violence. Unfortunately, both efforts allow the biases of researchers to distort their summaries of available evidence on the relationship between firearms and violence.
I
Under the Gun began as a grant to the principal authors from the [Page 956] National Institute of Justice. The authors were to do a comprehensive survey of existing literature on weapons, crime and violence and to prepare for the government an agenda for further empirical research on the topic. This two part task--survey what's there and tell us what's needed--is standard operating procedure for National Institute of Justice reviews of the state of the art in issues related to crime control. In this case, however, the Department of Justice was getting a late start on the topic: the 'war on crime' had gone on for nearly a decade and the federal government had invested hundreds of millions of dollars in research on crime and violence without supporting a single substantial investigation of the relationship between weapons (or firearms control) and violence.
The report of research activities from this project has now undergone two incarnations. The first was a final report to the federal government, issued in 1981, with a lengthy review of the existing literature and a much briefer proposed 'agenda of research.' [2] The present volume was adapted from that final report and we are told: 'There are only minor differences between the earlier and present versions . . ..' (p. xiii). One of those minor differences, however, is that the agenda for future research has been deleted from this book.
This de-emphasis of future research is astonishing because the major conclusion of the investigation is that 'despite the large number of studies that have been done, many critically important questions have not been adequately researched, and some of them have not been examined at all' (p. xi).
That practicing social scientists do not urge extensive research in such circumstances generates the sort of surprise one would encounter if a group of dairy farmers came out against cheese. In a world wearied by the self- serving claims of interest groups, this omission should have been an occasion for delight. However, given the sorry state of knowledge on this topic, and the importance of the issue to public policy, the failure of Wright and his colleagues to plead for more information is the central puzzle of the book and a problem of substantial importance to the field. I will indulge in speculation about the reasons for this extraordinary omission after describing some of the topics that the book does cover.
Under the Gun is divided into three major sections: gun ownership and use; the relationship between firearms and violence; and regimes of firearms control. The division of topics and the sequence of treatment are both conventional [3] and appropriate. The coverage of published literature is nearly complete, no small achievement in a field without a previously organized bibliography. [Page 957]
What the authors do with what they have found in the literature, however, is less impressive than the product of their bibliographic skill. 'Uneven' would be the most appropriate overall description of the quality of the critique. In general, the analysis of patterns of firearms ownership and use is stronger than the other two major sections. The analysis of strategies of gun control--how they might work and how they are flawed--is the analytic weak sibling of the effort.
The unevenness of the overall effort reflects the background of the authors in survey sociology and general social statistics. When discussing estimates of gun ownership and opinion poll data on reasons for ownership and attitudes toward gun control, Professor Wright and his associates are on familiar ground, and they cover it well. The discussion of estimates of civilian gun ownership in chapter 2 is the best treatment of that topic to date. The materials on public opinion toward gun control (chapter 11) and patterns of gun ownership (chapters 3 through 6) display a high degree of professional competence.
But the authors have no substantial background in the study of criminal violence or in legal studies and this, too, is evident. The review of evidence of the relationship between gun availability and the death rate from assault is sophomoric, unconstructive, and probably misleading. The discussion of firearms controls fails to outline the objectives of various strategies of control and the specific weaknesses of each approach. Instead, the authors fall back on a gestalt judgment that gun controls won't work because there are so many guns available and only a few are needed to supply criminal demands (pp. 189, 189 n.1, 320), a sweeping judgment that is probably wrong [4] and certainly a poor substitute for reasoned analysis.
The treatment of a series of studies seeking evidence about the effect of weapon dangerousness on the death rate from assault will serve as an example of the way in which violence research is discussed, since it is one of the most sustained critiques in the book and the work it discusses was written by my favorite author. The principal targets of Under the Gun's critique, two Chicago studies of fatal and nonfatal assaults, had reported:
(1) There was a substantial overlap in the circumstances, wound locations, and number of wounds inflicted between fatal and nonfatal assaults with guns and those committed with knives;
(2) Gun assaults were far more likely to lead to death than knife assaults and the large differences persisted when the comparisons were controlled for the number of wounds and the body location of the most serious wound; and
(3) Consistent with the 'instrumentality hypothesis' generated by the comparison of gun versus knife assaults, attacks with large caliber firearms were far more likely to cause death than woundings by small caliber [Page 958] guns that resulted in the same number of wounds to the same parts of the body. [5]
The authors of Under the Gun devote almost an entire chapter (chapter 11--'On the Matter of Criminal Motivations') to this evidence and find it wanting. Circumstantial evidence such as the motives of homicide and the frequent involvement of alcohol in killings is dismissed as inconclusive on the issue of whether attacks that cause death are often ambiguously motivated. The similar demographic profiles of fatal and nonfatal attackers is not regarded as evidence that the two groups 'are similar in any respect relevant to hypotheses about underlying motivations' (p. 196). The same conclusion apparently was applied to the similarities between victim groups.
The authors never address the possibility that chance elements select a subsample of fatalities from the universe of those assaulted. The crucial fact that most gun killings, like most nonfatal assaults, involve only one wound is rejected as evidence against a single-minded intent to kill in favor of the conclusion that what distinguishes the hundreds of one-shot killings from thousands of one-shot nonfatal woundings in the same body location by the same sort of people is 'a level of marksmanship that one would probably not expect under conditions of outrage and duress' (p. 196). This conclusion is reached without any evidence from the extensive literature on criminal violence.
Not that the authors lack knowledge they think relevant to this matter. A long textual footnote discusses Professor Wright's career in the butchering of game deer, noting that none of the carcasses he dressed 'was taken with one and only one shot. . . . That a much higher proportion of murderers, armed with much less impressive weaponry, kill with a single shot might therefore cause us to wonder just how ambiguous the underlying motives are' (p. 196 n.5).
That the authors neither consider the alternative hypothesis that many killers are randomly drawn from the larger pool of one-shot assaulters, nor discuss why 'determined' killers will often stop after one wounding when most guns have a multiple wounding capacity, is worrisome. I am also concerned by the fact that the sharp differences in death rates for large caliber versus small caliber gun assaults is not considered to be evidence that the objective dangerousness of a weapon has a significant influence on the death rate in assault. The authors briefly examine this data in another section of their chapter. They caution that this pattern could also be explained if more determined killers choose larger caliber guns (p. 203). The evidence from this weapon caliber study that the pattern holds true even when the [Page 959] attacker probably did not choose the weapon is ignored. [6]
What makes the critique on criminal motivation instructive, however, is neither the bias of the treatment nor the eminent target of the criticism. Incomplete renditions of available evidence abound in this book, [7] and the literary output of my heedless youth is frequently roasted. [8] Chapter 10, however, may be of special value in unraveling the mystery of the authors' missing agenda of research. The parable of the butchered deer carcass mentioned above seems the closest the authors can come to a personal background in research on criminal violence. Instead of grounding their discussion in a coherent vision of violent assault, they put forward rival hypotheses to each strand of circumstantial evidence individually, conclude that none is strict proof of instrumentality effects by itself, and assume that the cumulative impact of multiple strands of evidence is no more persuasive than any of the individual strands.
When the time comes to recommend future research, the authors are prisoners of their own standards. In the report to the federal government, no further investigation of these critical issues is proposed. In this book, the authors leave the impression that nothing is known on the question of what difference guns make and nothing can be done to increase knowledge. The tone, and the level of denial, remind one of the Tobacco Institute's valiant struggle against premature conclusions on the relationship between cigarettes and lung cancer. [9]
If nothing is known about the cost of gun use in violent crime, how can reasoned decisions be made about the benefits and costs of various schemes to regulate guns? The only excuse for failure to persist in measuring the consequences of gun use is determining that gun laws cannot work or will be so costly as to outweigh any attainable benefits. As long as gun effects cannot be measured, this is the only firm conclusion to which any summary statement can come. Given the pressure to reach some conclusion after all that effort, one might expect a temptation toward nihilistic conclusions on such matters that has nothing to do with general political ideology. The temptation to treat gun control [Page 960] proposals as the Emperor's new clothes might be reinforced by the almost automatic endorsement of gun control that is current among political liberals and most of the social science community. A revisionist view has news value of no small dimensions. Is this what has happened here?
For whatever reason, the analysis of specific strategies of gun control in this volume is far less organized and less rigorous than the analysis in other parts of the book. Sandwiched between a good treatment of public opinion polls and a competent summary of existing evaluation strategies, chapter 12 'discusses the kinds of laws that have been passed and, to some extent, their intended (or hoped for) effects' (p. 244). The chapter does categorize existing state and local regulation to some extent, but there is no analysis of what intended impact such regimes as owner licensing, regulation, prohibition of ownership of guns by high risk groups, waiting periods, 'place and manner laws,' transfer notice and other regulatory approaches have alone or in combination.
The detailed analysis of specific gun control strategies is, instead, pre-empted by the book's endorsement of what I shall call a 'needle-in-the- haystack' perspective, the attempt to minimize the potential of gun regulation to reduce firearm-related violence because there are so many guns now in civilian hands and only a small proportion of those are involved in criminal violence:
It may be taken as self-evident that something in excess of 99% of all privately owned firearms are never involved in any sort of criminal act, and it is certainly possible . . . that the criminally abused 1% would be the last guns touched by any sort of restrictive weapons policy. [P. 189 n.1.]
The significance of these statistics for gun control prospects is said to be that 'the existing stock [of guns] is adequate to supply all conceivable criminal purposes for at least the entire next century, even if the world-wide manufacture of new guns were halted today' (p. 320).
Three points should be made about these statistics and the inferences that are drawn from them. First, the interpretations made by the authors in the foregoing are both inconsistent and wrong. It is simply not true that ninety- nine percent of all guns are never involved in crime. Instead, fewer than one percent of all guns are involved in criminal misuse in any given year, just as less than one percent of the American population dies of heart disease in a single year. But that doesn't make either heart disease or gun related crimes a small problem. The career risk of guns being misused is much greater, though not as great as the second quote from Under the Gun would suggest. Probably more than ten percent of all handguns are used in crime or [Page 961] serious violence [10]--usually within a decade of first sale--but a much smaller fraction of all long guns are so used.
There is also ample evidence that the existing stock of civilian guns is nowhere near a century's supply for criminal users. Only by ignoring existing data on the firearms used in crime and assuming that guns owned by anyone are freely available to all who would misuse them can the extravagant claims about 'a century's supply' be given any semblance of a conceptual foundation.
Although this book's conclusions from statistics on civilian ownership are fantastic, any balanced view of the potential of different gun control options must start with data on current ownership and use of guns and how these patterns might be affected by various policy options. That is why detailed knowledge of gun ownership, the determinants of availability to potential gun misusers, and the effect of controls on gun availability in the middle and long run are vital in assessing the prospects of different proposals for reducing gun violence. Information on existing gun ownership and gun use in crime, however, is only a starting point for analysis rather than a short cut to a general conclusion. This book's attempt to use that first step as the basis for sweeping policy conclusions is an underachievement of substantial importance. Cautious to a significant fault is analysis of so much previous work, the book lurches to unjustified generalizations just when painstaking specificity is needed.
II
While parts of Under the Gun can be criticized for failure to draw inferences when the facts support them, Firearms and Violence frequently deserves Olympic medals for jumping to conclusions. This volume is a collection of essays on specific gun-related topics. The result is a contrast, in two respects, to most essay collections in which different authors write on the same topic. First, the chapters were commissioned by the editor so that the topics covered converge into a more coordinated sequence than edited volumes usually produce, with sections on public opinion, gun ownership and crime, handgun-only control, gun law impact, self-defense, the second amendment, and the social dimensions of gun ownership. While a substantial amount of overlap among the essays in this volume with regard to content and studies cited remains, things could have been much worse. Only the two 'overview' essays at the beginning of the volume (neither very helpful), John Kaplan's intelligent and free-standing foreword, and the editor's attempt at synthesis in the conclusion fail to address in detail specific topics not covered by other chapters. [Page 962]
The second distinction between this volume and most edited collections concerns the editorial slant of the contributions. A standard feature of such collections is that they fail to speak from a single perspective. Not here. With few exceptions, the essays collected in these pages reflect the wonderful world of Don B. Kates, Jr., a world in which guns do not contribute to the cost of interpersonal violence (and indeed, they prevent violent crime), handguns are safer than long guns, and any attempts to reduce gun violence through regulation of gun availability are simultaneously ineffectual and totalitarian.
This world view can fairly be called 'wonderful' because all of the policy choices to be made are easy and painless. If there are, in fact, reasons to worry about guns and their control, the reader must search elsewhere to find them. The statistical brief that supports this position includes a great deal of new material and a substantial amount of creativity. In contrast to Under the Gun, this collection of materials contains original scholarship on a variety of topics, and specific analysis of different types of firearms controls. In fact, there is more new statistical analysis in this volume than can be found in any recent mainstream publication on the topic. Moreover, these authors frequently display knowledge about firearms, awareness of existing studies, and a resourcefulness in the use of statistics that is truly impressive.
That's the good news. The bad news is that two plus two do not always equal four when statistics are manipulated in this volume, undermining the credibility of the effort and again demonstrating the frailty of partisan research. Space does not permit a complete inventory of statistical infelicities, but let us consider a few highlights.
Professor Gary Kleck tests the hypothesis that new guns have a disproportionate impact on rates of gun violence by comparing trends in gun introduction in a given year with that same year's homicide total. The problem is that this tests the impact of December's gun introductions on the previous January's gun killings. Testing the impact of this year's gun introductions on next year's homicide, seemingly a rather natural strategy, is rejected because those guns aren't new enough (p. 123), despite data that show new guns to have a significant impact for many years. [11]
A second example concerns the needle-in-the-haystack argument previously discussed. The reader may recall my complaints that Wright, Rossi, and Daly understate the eventual dangerousness of civilian handguns by at least a factor of ten. Kates makes this look like small beer, arguing for a ratio of handgun criminals to total handguns of 1 to 600 and a ratio of handgun murderers to handguns of 1 to 5,400 (p. 528 n.9). If Under the Gun speaks of a 'century's supply' of crime guns, Kates appears to up the ante to a millenium's worth. [Page 963]
James Wright's (one of the co-authors of Under the Gun) contribution to Kates' volume concerning self-defense (chapter 11), asserts that 'there are at least as many crimes 'thwarted' by the victim actually shooting at the offender as there are offenders who are apprehended and imprisoned for their offense' (p. 314). In fact, Wright provides no data that supports this hypothesis and his own calculations show a rate of imprisonment at least four- and-one-half times the most optimistic prevention-by-shooting estimate he himself cited only two pages earlier (p. 312).
Chapter 7 on 'Handgun Only Gun-Control' by Professor Kleck provides a panoramic view of both the creativity and the limits of the statistical adventures in this book. Here, Kleck's argument is that restricted availability of handguns could result in a higher death rate from attacks if more deadly weapons like rifles and shotguns were substituted. Kleck presents no data on the degree to which rifles and shotguns would or do displace scarce handguns in the United States or in any other nation. He treats a guess in the literature that the displacement to long guns would be no more than one-third as the equivalent to a finding that it would be a third, a logical slip of a sort frequently encountered in these materials (p. 176). He presents no data on the death rate from various types of assault with long weapons, saying, 'it would be pointless to compare actual observed assault fatality rates of handguns and long guns to determine relative deadliness, even if adequate data were available for such an effort, since the fatality rates are not just the result of the deadliness of the weapons themselves' (p. 177). Instead, the author uses a concept of the relative 'stopping power' of guns which 'should clearly be positively correlated with deadliness' (p. 177).
As refreshing as it is to see instrumentality effects acknowledged by one of their critics, the superiority of 'stopping power' over the actual death rate from attacks escapes me. The use of this method results in estimates that some guns, such as the twelve-gauge shotgun with double-ought cartridges, would produce fatalities in head and chest wounds about eighteen times as often as . 22-caliber rifles, or approximately three deaths for every individual attacked. [12]
This problem is compounded by the failure to provide any data on the profile of rifles and shotguns used in assaults. The author uses one very biased sample of long guns used in homicide [13] to project the distribution [Page 964] of long guns in all assaults, despite the fact that Kleck's estimates of deadliness would argue that .22-caliber rifles are underestimated relative to twelve-gauge shotguns by a factor of eighteen, since they will result in a much smaller number of attacks leading to fatality. These are not small problems.
This excursion into pseudo-mathematics has diagnostic value, for it is now necessary to ask about the mind-set that can lead to death rate estimates of three per wounding and a comparison of the deadlines of different guns that deliberately avoids studying the use of those guns in actual attacks. It must require a potent combination of wishful thinking, reciprocal noncriticism by peers, and absence of self-examination for this kind of catastrophic error to march into the public debate on guns and gun control.
In such circumstances, I am at a loss to comment on the actual value of scholarship in this book outside areas of my personal knowledge. There is, for instance, some interesting discussion of the common law background on civilian firearms ownership that inevitably suffers, in my mind, from the low credibility of the work that I can more certainly evaluate. The editor's pretensions to nonpartisanship are another minor irritant ('the papers in this volume do not speak from either of the conflicting viewpoints that have . . . so dominated the American gun control debate' (p. 524).)
The snide commentary, bunker mentality, and penchant for quotation out of context in many of the papers do not rankle as much in this collection, however, as the patronizing tone of Under the Gun. Most of the contributors to the Kates collection seem passionately sincere, but it was Oscar Wilde who first observed that excess quantities of sincerity can prove fatal.
There is and should be room for this kind of polemic in the dialogue on guns in the United States. It is the absence of a stabilizing core of solid research and analysis that is the occasion for sadness and concern.
* Llewellyn Professor of Jurisprudence and Director of the Center for Studies in Criminal Justice, the University of Chicago Law School; Visiting Professor of Law, University of California, School of Law, Boalt Hall. B.A. 1963, Wayne State University; J.D. 1967, University of Chicago Law School.--Ed.
[1]. G. NEWTON & F. ZIMRING, FIREARMS AND VIOLENCE IN AMERICAN LIFE, TASK REPORT NO. 7 TO THE NATIONAL COMMISSION ON THE CAUSES AND PREVENTION OF VIOLENCE (1969).
[2]. J. WRIGHT & P. ROSSI, WEAPONS, CRIME AND VIOLENCE IN AMERICA (1981).
[3]. See G. NEWTON & F. ZIMRING, supra note 1, at xix-xx.
[4]. See text at note 10 infra.
[5]. See Zimring, Is Gun Control Likely to Reduce Violent Killings?, 35 U. CHI. L. REV. 721 (1968); Zimring, The Medium is the Message: Firearm Caliber as a Determinant of Death From Assault, 1 J. LEGAL STUD. 97 (1972) [hereinafter cited as The Medium is the Message].
[6]. The comparisons here were limited to domestic assaults, where the shooter presumably did not have that use in mind when selecting the caliber of weapon in the more than four out of five cases where the household owned only one handgun, and gun assaults by women who usually did not buy the guns available in the household. See The Medium is the Message, supra note 5, at 101-03.
[7]. Worthy of special note are the argument in chapter 5 that fear of crime did not motivate the expansion of handgun ownership in the 1960's since it may not have been the motive for handgun acquisition in the 1970's, and the analysis of self-defense handguns in chapter 7 that fails to mention studies of communities where more persons are killed by handgun accidents than by home- intruding strangers.
[8]. See, e.g., the entries under 'Zimring' at p. 342 and the textual references there listed.
[9]. See Smoking and Health 1964-1979. The Continuing Controversy: Hearings on H.R. 1824 Before the Subcomm. on Health and the Environment of the Comm. on Energy and Commerce of the House of Representatives, 98th Congress, 1st Sess. 544 (1983) (submitted by the Tobacco Institute).
[10]. See Cook, Guns and Crime: The Peril of Long Division, 81 J. POLY. ANALYSIS & MGMT. 120 (1981).
[11]. Zimring, Street Crime and New Guns. Some Implications for Firearms Control, 4 J. CRIM. JUST. 95 (1976).
[12]. Brian Gluss first drew my attention to the mad mathematics of relative stopping power. The death rate from .22-caliber, single-wound attacks with a head or chest injury is reported at 16% in The Medium is the Message, supra note 5, at 104. Multiplying this death rate by 18 (the difference in relative stopping power) provided an estimated death rate of 288% for single- wound attacks with 12-gauge shotguns that result in head or chest wounds. The death rate from multiple wound .22-caliber attacks in that study was reported at 28%. Multiplying this estimate by 18 produces an estimated death rate of 504%, or five deaths for every individual wounded.
[13]. The problem in the sample used is that less dangerous low caliber rifles will less often be represented in weapons causing death than in those causing nonfatal injury. This bias is compounded by the fact that the police will often not know whether a killing was caused with a .22-caliber rifle or a .22 handgun unless the weapon is recovered. Shotgun killings are easy to identify even if the specific weapon cannot be found.