Harvard - Would Banning Firearms Reduce Murders and Suicide

WOULD BANNING FIREARMS REDUCE
MURDER AND SUICIDE?
A REVIEW OF INTERNATIONAL AND
SOME DOMESTIC EVIDENCE
DON B. KATES* AND GARY MAUSER**
INTRODUCTION ............................................................650
I. VIOLENCE: THE DECISIVENESS OF
SOCIAL FACTORS ...................................................660
II. ASKING THE WRONG QUESTION ..........................662
III. DO ORDINARY PEOPLE MURDER?........................665
IV. MORE GUNS, LESS CRIME?....................................670
V. GEOGRAPHIC, HISTORICAL AND DEMOGRAPHIC
PATTERNS ...............................................................673
A. Demographic Patterns ..................................676
B. Macro‐historical Evidence: From the
Middle Ages to the 20th Century .................678
C. Later and More Specific Macro‐Historical
Evidence..........................................................684
D. Geographic Patterns within Nations ..........685

* Don B. Kates (LL.B., Yale, 1966) is an American criminologist and constitutional
lawyer associated with the Pacific Research Institute, San Francisco. He may be con‐
tacted at [email protected]; 360‐666‐2688; 22608 N.E. 269th Ave., Battle Ground,
WA 98604.
** Gary Mauser (Ph.D., University of California, Irvine, 1970) is a Canadian crimi‐
nologist and university professor at Simon Fraser University, Burnaby, BC Canada.
He may be contacted at www.garymauser.net, [email protected], and 604‐291‐3652.
We gratefully acknowledge the generous contributions of Professor Thomas B. Cole
(University of North Carolina at Chapel Hill, Social Medicine and Epidemiology);
Chief Superintendent Colin Greenwood (West Yorkshire Constabulary, ret.); C.B.
Kates; Abigail Kohn (University of Sydney, Law); David B. Kopel (Independence
Institute); Professor Timothy D. Lytton (Albany Law School); Professor William
Alex Pridemore (University of Oklahoma, Sociology); Professor Randolph Roth
(Ohio State University, History); Professor Thomas Velk (McGill University, Eco‐
nomics and Chairman of the North American Studies Program); Professor Robert
Weisberg (Stanford Law School); and John Whitley (University of Adelaide, Eco‐
nomics). Any merits of this paper reflect their advice and contributions; errors are
entirely ours.
650 Harvard Journal of Law & Public Policy [Vol. 30
E. Geographic Comparisons: European
Gun Ownership and Murder Rates ..........687
F. Geographic Comparisons: Gun‐Ownership
and Suicide Rates ...........................................690
CONCLUSION ................................................................693

INTRODUCTION
International evidence and comparisons have long been offered
as proof of the mantra that more guns mean more deaths and that
fewer guns, therefore, mean fewer deaths.1 Unfortunately, such
discussions are all too often been afflicted by misconceptions and
factual error and focus on comparisons that are unrepresentative.
It may be useful to begin with a few examples. There is a com‐
pound assertion that (a) guns are uniquely available in the United
States compared with other modern developed nations, which is
why (b) the United States has by far the highest murder rate.
Though these assertions have been endlessly repeated, statement
(b) is, in fact, false and statement (a) is substantially so.
Since at least 1965, the false assertion that the United States has
the industrialized world’s highest murder rate has been an artifact
of politically motivated Soviet minimization designed to hide the
true homicide rates.2 Since well before that date, the Soviet Union

1. See, e.g., JOHN GODWIN, MURDER USA: THE WAYS WE KILL EACH OTHER 281
(1978) (“Areas with the highest proportion of gun owners also boast the highest
homicide ratios; those with the fewest gun owners have the lowest.”); N. PETE
SHIELDS, GUNS DON’T DIE, PEOPLE DO 64 (1981) (quoting and endorsing an English
academic’s remark: “We cannot help but believe that America ought to share the
basic premise of our gun legislation—that the availability of firearms breeds vio‐
lence.”); Janice Somerville, Gun Control as Immunization, AM. MED. NEWS, Jan. 3,
1994, at 9 (quoting public health activist Katherine Christoffel, M.D.: “Guns are a
virus that must be eradicated . . . . Get rid of the guns, get rid of the bullets, and you
get rid of the deaths.”); Deane Calhoun, From Controversy to Prevention: Building Ef‐
fective Firearm Policies, INJ. PROTECTION NETWORK NEWSL., Winter 1989–90, at 17
(“[G]uns are not just an inanimate object [sic], but in fact are a social ill.”); see also
WENDY CUKIER & VICTOR W. SIDEL, THE GLOBAL GUN EPIDEMIC: FROM SATURDAY
NIGHT SPECIALS TO AK‐47S (2006); Susan Baker, Without Guns, Do People Kill People?
75 AM. J. PUB. HEALTH 587 (1985); Paul Cotton, Gun‐Associated Violence Increasingly
Viewed as Public Health Challenge, 267 J. AM. MED. ASS’N 1171 (1992); Diane Schetky,
Children and Handguns: A Public Health Concern, 139 AM. J. DIS. CHILD. 229, 230 (1985);
Lois A. Fingerhut & Joel C. Kleinman, International and Interstate Comparisons of
Homicides Among Young Males, 263 J. AM. MED. ASS’N 3292, 3295 (1990).
2. See William Alex Pridemore, Using Newly Available Homicide Data to Debunk Two
Myths About Violence in an International Context: A Research Note, 5 HOMICIDE STUD.
267 (2001).
No. 2] Would Banning Firearms Reduce Murder and Suicide? 651
possessed extremely stringent gun controls3 that were effectuated
by a police state apparatus providing stringent enforcement.4 So
successful was that regime that few Russian civilians now have
firearms and very few murders involve them.5 Yet, manifest suc‐
cess in keeping its people disarmed did not prevent the Soviet
Union from having far and away the highest murder rate in the
developed world.6 In the 1960s and early 1970s, the gun‐less So‐
viet Union’s murder rates paralleled or generally exceeded those
of gun‐ridden America. While American rates stabilized and then
steeply declined, however, Russian murder increased so drasti‐
cally that by the early 1990s the Russian rate was three times
higher than that of the United States. Between 1998‐2004 (the lat‐
est figure available for Russia), Russian murder rates were nearly
four times higher than American rates. Similar murder rates also
characterize the Ukraine, Estonia, Latvia, Lithuania, and various
other now‐independent European nations of the former U.S.S.R.7
Thus, in the United States and the former Soviet Union transition‐
ing into current‐day Russia, “homicide results suggest that where

3. See GEORGE NEWTON & FRANKLIN ZIMRING, FIREARMS AND VIOLENCE IN
AMERICAN LIFE: A STAFF REPORT SUBMITTED TO THE NATIONAL COMMISSION ON THE
CAUSES AND PREVENTION OF VIOLENCE 119 & n.3 (1970).
4. Russian law flatly prohibits civilian possession of handguns and limits long
guns to licensed hunters. Id. For more on the stringency of enforcement, see Ray‐
mond Kessler, Gun Control and Political Power, 5 LAW & POL’Y Q. 381, 389 (1983), and
Randy E. Barnett & Don B. Kates, Under Fire: The New Consensus on the Second
Amendment, 45 EMORY L. J. 1139, 1239 (1996) (noting an unusual further element of
Soviet gun policy: the Soviet Army adopted unique firearm calibers so that, even if
its soldiers could not be prevented from returning with foreign gun souvenirs from
foreign wars, ammunition for them would be unavailable in the Soviet Union).
5. See Pridemore, supra note 2, at 271.
6. Russian homicide data given in this article (for years 1965–99) were kindly sup‐
plied us by Professor Pridemore from his research in Russian ministry sources (on
file with authors). See also infra Table 1 (reporting Russian homicide data for 2002).
7. The highest U.S. homicide rate ever reported was 10.5 per 100,000 in 1980. See
Jeffery A. Miron, Violence, Guns, and Drugs: A Cross‐Country Analysis, 44 J.L. & ECON.
615, 624–25 tbl.1 (2001). As of 2001, the rate was below 6. Id. The latest rates available
for the Ukraine, Belarus, and other former Soviet nations in Europe come from the
mid‐1990s, when all were well above 10 and most were 50% to 150% higher. Id.
Note that the U.S. rates given above are rates reported by the FBI. There are two
different sources of U.S. murder rates. The FBI murder data is based on reports it
obtains from police agencies throughout the nation. These data are significantly less
complete than the alternative (used in this article unless otherwise explicitly stated)
rates of the U.S. Public Health Service, which are derived from data collected from
medical examiners’ offices nationwide. Though the latter data are more comprehen‐
sive, and the Public Health Service murder rate is slightly higher, they have the dis‐
advantage of being slower to appear than the FBI homicide data.
guns are scarce other weapons are substituted in killings.”8 While
American gun ownership is quite high, Table 1 shows many other
developed nations (e.g., Norway, Finland, Germany, France,
Denmark) with high rates of gun ownership. These countries,
however, have murder rates as low or lower than many devel‐
oped nations in which gun ownership is much rarer. For example,
Luxembourg, where handguns are totally banned and ownership
of any kind of gun is minimal, had a murder rate nine times
higher than Germany in 2002.9
Table 1: European Gun Ownership and Murder Rates
(rates given are per 100,000 people and in descending order)
Nation Murder Rate Rate of Gun Ownership

Russia 20.54 [2002] 4,000
Luxembourg 9.01 [2002] c. 0
Hungary 2.22 [2003] 2,000
Finland 1.98 [2004] 39,000
Sweden 1.87 [2001] 24,000
Poland 1.79 [2003] 1,500
France 1.65 [2003] 30,000
Denmark 1.21 [2003] 19,000
Greece 1.12 [2003] 11,000
Switzerland 0.99 [2003] 16,000
Germany 0.93 [2003] 30,000
Norway 0.81 [2001] 36,000
Austria 0.80 [2002] 17,000
Notes: This table covers all the Continental European nations for which
the two data sets given are both available. In every case, we have given
the homicide data for 2003 or the closest year thereto because that is the
year of the publication from which the gun ownership data are taken. Gun
ownership data comes from GRADUATE INSTITUTE OF INTERNATIONAL
STUDIES, SMALL ARMS SURVEY 64 tbl.2.2, 65 tbl.2.3 (2003).
The homicide rate data comes from an annually published report,
CANADIAN CENTRE FOR JUSTICE STATISTICS, HOMICIDE IN CANADA,
JURISTAT, for the years 2001–2004. Each year’s report gives homicide sta‐
tistics for a dozen or so foreign nations in a section labeled “Homicide
Rates for Selected Countries.” This section of the reports gives no explana‐

8. GARY KLECK, TARGETING GUNS: FIREARMS AND THEIR CONTROL 20 (1997) (dis‐
cussing patterns revealed by studies in the United States).
9. Our assertions as to the legality of handguns are based on COMM’N ON CRIME
PREVENTION & CRIM. JUSTICE, U.N. ECON. & SOC. COUNCIL, UNITED NATIONS
INTERNATIONAL STUDY ON FIREARMS REGULATION 26, tbl. 2‐1 (1997 draft).
tion of why it selects the various nations whose homicide statistics it cov‐
ers. Also without explanation, the nations covered differ from year to
year. Thus, for instance, murder statistics for Germany and Hungary are
given in all four of the pamphlets (2001, 2002, 2003, 2004), for Russia in
three years (2001, 2002, and 2004), for France in two years (2001 and 2003),
and for Norway and Sweden in only one year (2001).

The same pattern appears when comparisons of violence to
gun ownership are made within nations. Indeed, “data on fire‐
arms ownership by constabulary area in England,” like data
from the United States, show “a negative correlation,”10 that is,
“where firearms are most dense violent crime rates are lowest,
and where guns are least dense violent crime rates are high‐
est.”11 Many different data sets from various kinds of sources
are summarized as follows by the leading text:
[T]here is no consistent significant positive association be‐
tween gun ownership levels and violence rates: across (1)
time within the United States, (2) U.S. cities, (3) counties
within Illinois, (4) country‐sized areas like England, U.S.
states, (5) regions of the United States, (6) nations, or (7)
population subgroups . . . .12
A second misconception about the relationship between fire‐
arms and violence attributes Europe’s generally low homicide

10. JOYCE LEE MALCOLM, GUNS AND VIOLENCE: THE ENGLISH EXPERIENCE 204
(2002).
11. Hans Toch & Alan J. Lizotte, Research and Policy: The Case for Gun Control, in
PSYCHOLOGY & SOCIAL POLICY 223, 232 (Peter Suedfeld & Philip E. Tetlock eds.,
1992); see also id. at 234 & n.10 (“[T]he fact that national patterns show little violent
crime where guns are most dense implies that guns do not elicit aggression in any
meaningful way. . . . Quite the contrary, these findings suggest that high saturations
of guns in places, or something correlated with that condition, inhibit illegal aggres‐
sion.”).
Approaching the matter from a different direction, the earliest data (nineteenth
century on) reveals that the American jurisdictions with the most stringent gun con‐
trols are in general the ones with the highest murder rates. Conversely, American
states with homicide rates as low as Western Europe’s have high gun ownership,
and impose no controls designed to deny guns to law‐abiding, responsible adults.
Many possible reasons may be offered for these two facts, but none suggests that
gun control reduces murder.
For examination of a wide variety of studies finding little evidence in support of
the efficacy of gun controls in reducing violence, see JAMES B. JACOBS, CAN GUN
CONTROL WORK? 111–20 (2002); KLECK, supra note 8, at 351–77; JOHN R. LOTT, JR.,
MORE GUNS, LESS CRIME: UNDERSTANDING CRIME AND GUN CONTROL LAWS 19–20
(1998); JAMES D. WRIGHT ET AL., UNDER THE GUN: WEAPONS, CRIME AND VIOLENCE
IN AMERICA 307–08 (1983); Matthew R. DeZee, Gun Control Legislation: Impact and
Ideology, 5 LAW & POL’Y Q. 367, 369–71 (1983).
12. KLECK, supra note 8, at 22‐23.
rates to stringent gun control. That attribution cannot be accu‐
rate since murder in Europe was at an all‐time low before the
gun controls were introduced.13 For instance, virtually the only
English gun control during the nineteenth and early twentieth
centuries was the practice that police patrolled without guns.
During this period gun control prevailed far less in England or
Europe than in certain American states which nevertheless
had—and continue to have—murder rates that were and are
comparatively very high.14
In this connection, two recent studies are pertinent. In 2004,
the U.S. National Academy of Sciences released its evaluation
from a review of 253 journal articles, 99 books, 43 government
publications, and some original empirical research. It failed to
identify any gun control that had reduced violent crime, sui‐
cide, or gun accidents.15 The same conclusion was reached in
2003 by the U.S. Centers for Disease Control’s review of then‐
extant studies.16
Stringent gun controls were not adopted in England and
Western Europe until after World War I. Consistent with the
outcomes of the recent American studies just mentioned, these
strict controls did not stem the general trend of ever‐growing
violent crime throughout the post‐WWII industrialized world
including the United States and Russia. Professor Malcolm’s
study of English gun law and violent crime summarizes that

13. Barnett & Kates, supra note 4, at 138–42.
14. In the period between 1900 and 1935, Arkansas, Hawaii, Michigan, Missouri,
New Jersey, New York, North Carolina, Oregon, and South Carolina adopted laws
variously requiring a license to own or buy a handgun or banning handgun pur‐
chase altogether, and “Saturday Night Special”‐type bans existed in Tennessee, Ar‐
kansas, and various other Southern states. Don B. Kates, Jr., Toward a History of
Handgun Prohibition in the United States, in RESTRICTING HANDGUNS: THE LIBERAL
SKEPTICS SPEAK OUT 7, 14–15 (Don B. Kates, Jr. ed., 1979).
15. CHARLES F. WELLFORD ET AL., NAT’L RESEARCH COUNCIL, FIREARMS AND
VIOLENCE: A CRITICAL REVIEW 6–10 (2004). It is perhaps not amiss to note that the
review panel, which was set up during the Clinton Administration, was composed
almost entirely of scholars who, to the extent their views were publicly known be‐
fore their appointments, favored gun control.
16. Task Force on Community Preventitive Servs., Ctrs. for Disease Control, First
Reports Evaluating the Effectiveness of Strategies for Preventing Violence: Firearms Laws,
52 MORTALITY & MORBIDITY WKLY. REP. (RR‐14 RECOMMENDATIONS & REP.) 11, 16
(2003), available at http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5214a2.htm.
The CDC is vehemently anti‐gun and interpreted its results to show not that the
“more guns equal more death” mantra is erroneous, but only that the scores of stud‐
ies it reviewed were inconclusively done.
nation’s nineteenth and twentieth century experience as fol‐

lows:
The peacefulness England used to enjoy was not the result of
strict gun laws. When it had no firearms restrictions [nine‐
teenth and early twentieth century] England had little vio‐
lent crime, while the present extraordinarily stringent gun
controls have not stopped the increase in violence or even
the increase in armed violence.17
Armed crime, never a problem in England, has now be‐
come one. Handguns are banned but the Kingdom has mil‐
lions of illegal firearms. Criminals have no trouble finding
them and exhibit a new willingness to use them. In the dec‐
ade after 1957, the use of guns in serious crime increased a
hundredfold.18
In the late 1990s, England moved from stringent controls to a
complete ban of all handguns and many types of long guns.
Hundreds of thousands of guns were confiscated from those
owners law‐abiding enough to turn them in to authorities.
Without suggesting this caused violence, the ban’s ineffective‐
ness was such that by the year 2000 violent crime had so in‐
creased that England and Wales had Europe’s highest violent
crime rate, far surpassing even the United States.19 Today, Eng‐
lish news media headline violence in terms redolent of the
doleful, melodramatic language that for so long characterized
American news reports.20 One aspect of England’s recent ex‐

17. MALCOLM, supra note 10, at 219.
18. Id. at 209.
19. See Esther Bouten et al., Criminal Victimization in Seventeen Industrialized Coun‐
tries, in CRIME VICTIMIZATION IN COMPARATIVE PERSPECTIVE: RESULTS FROM THE
INTERNATIONAL CRIME VICTIMS SURVEY, 1989–2000 at 13, 15–16 (Paul Nieuwbeerta
ed., 2002). The surveys involved were conducted under the auspices of the govern‐
ments of each nation and the general supervision of the University of Leiden and
the Dutch Ministry of Justice.
20. See, e.g., Gun Crime Growing “Like Cancer,” BBC NEWS, May 21, 2003,
http://news.bbc.co.uk/1/hi/England/3043701.stm; David Bamber, Gun Crime Trebles
as Weapons and Drugs Flood British Cities, TELEGRAPH ONLINE (London), Feb. 27, 2002,
http://www.telegraoh.co.uk/news/main.jhtml?xml=/news/2002/02/24/nguns24.xml;
Jason Bennetto, Firearms Amnesty to Tackle Surge in Gun Crime, INDEPENDENT (Lon‐
don), Dec. 27, 2002, at 1; Ian Burrell, Police Move to Tackle Huge Rise in Gun Crime,
INDEPENDENT (London), Jan. 15, 2001, at 3; Daniel Foggo & Carl Fellstrom, “We Are
Reeling with the Murders, We Are in a Crisis with Major Crime,” SUNDAY TELEGRAPH
(London), Mar. 13, 2005, at 4; Johann Hari, The British Become Trigger Happy, NEW
STATESMAN (London), Nov. 5, 2001, at 35; Philip Johnston, Gun Crime Rises Despite
Dunblane Pistol Ban, DAILY TELEGRAPH (London), Jul. 17, 2001, at 05; David Leppard
& Rachel Dobson, Murder Rate Soars to Highest for a Century, SUNDAY TIMES (Lon‐

perience deserves note, given how often and favorably advo‐

cates have compared English gun policy to its American coun‐
terpart over the past 35 years.21 A generally unstated issue in
this notoriously emotional debate was the effect of the Warren
Court and later restrictions on police powers on American gun
policy. Critics of these decisions pointed to soaring American
crime rates and argued simplistically that such decisions
caused, or at least hampered, police in suppressing crime. But
to some supporters of these judicial decisions, the example of
England argued that the solution to crime was to restrict guns,
not civil liberties. To gun control advocates, England, the cradle
of our liberties, was a nation made so peaceful by strict gun
control that its police did not even need to carry guns. The
United States, it was argued, could attain such a desirable
situation by radically reducing gun ownership, preferably by
banning and confiscating handguns.
The results discussed earlier contradict those expectations. On
the one hand, despite constant and substantially increasing gun
ownership, the United States saw progressive and dramatic re‐
ductions in criminal violence in the 1990s. On the other hand, the
same time period in the United Kingdom saw a constant and
dramatic increase in violent crime to which England’s response
was ever‐more drastic gun control including, eventually, banning
and confiscating all handguns and many types of long guns.22
Nevertheless, criminal violence rampantly increased so that by
2000 England surpassed the United States to become one of the
developed world’s most violence‐ridden nations.

don), Oct. 13, 2002, at 1; Adam Mitchell, Gun Killings Double as Police Claim Progress,
DAILY TELEGRAPH (London), Aug. 17, 2001, at 13; John Steele, Police Fear a New Crime
Wave as School‐Age Muggers Graduate to Guns, DAILY TELEGRAPH (London), Jan. 3,
2002, at 04; Jon Ungoed‐Thomas, Killings Rise as 3m Illegal Guns Flood Britain, SUN‐
DAY TIMES (London), Jan. 16, 2000; Peter Woolrich, Britain’s Tough Gun Control Laws
Termed Total Failure: Land of Hope and Gunrunning, PUNCH MAG., May 3, 2000.
21. See, e.g., CARL BAKAL, THE RIGHT TO BEAR ARMS 10–11, 31, 279 (1966); RAMSEY
CLARK, CRIME IN AMERICA 104–05, 109 (1970); AMITAI ETZIONI & RICHARD REMP,
TECHNOLOGICAL SHORTCUTS TO SOCIAL CHANGE 136 (1973); Natl Coalition to Ban
Handguns, A Shooting Gallery Called America (undated, unpaginated pamphlet);
SHIELDS, supra note 1, at 63–64; Irwin Bloch, Gun Control Would Reduce Crime, re‐
printed in Would Gun Control Reduce Crime 197 (David Bender ed., 1989); Robert S.
Drinan, Banning Handguns Would Reduce Crime, reprinted in GUNS & CRIME 45–46
(Tarara Roleff ed., 1999).
22. MALCOLM, supra note 10, at 164–216. We should clarify that the twin trends
toward more violent crime and more gun control began long before the 1990s. See id.
To conserve the resources of the inundated criminal justice

system, English police no longer investigate burglary and “mi‐
nor assaults.”23 As of 2006, if the police catch a mugger, robber,
or burglar, or other “minor” criminal in the act, the policy is to
release them with a warning rather than to arrest and prosecute
them.24 It used to be that English police vehemently opposed
the idea of armed policing. Today, ever more police are being
armed. Justifying the assignment of armed squads to block
roads and carry out random car searches, a police commander
asserts: “It is a massive deterrent to gunmen if they think that
there are going to be armed police.”25 How far is that from the
rationale on which 40 American states have enacted laws giv‐
ing qualified, trained citizens the right to carry concealed guns?
Indeed, news media editorials have appeared in England argu‐
ing that civilians should be allowed guns for defense.26 There is
currently a vigorous controversy over proposals (which the
Blair government first endorsed but now opposes) to amend
the law of self‐defense to protect victims from prosecution for
using deadly force against burglars.27
The divergence between the United States and the British
Commonwealth became especially pronounced during the

23. Daniel Foggo, Donʹt Bother About Burglary, Police Told, SUNDAY TELEGRAPH
(LONDON), Jan. 12, 2003, at 1 (“Police have been ordered not to bother investigating
crimes such as burglary, vandalism and assaults unless evidence pointing to the
culprits is easily available, The Sunday Telegraph can reveal. Under new guidelines,
officers have been informed that only ʺseriousʺ crimes, such as murder, rape or so‐
called hate crimes, should be investigated as a matter of course. In all other cases,
unless there is immediate and compelling evidence, such as fingerprints or DNA
material, the crime will be listed for no further action.”).
24. Steve Doughty, Let Burglars Off With Caution Police Told, DAILY MAIL (London),
Apr. 3, 2006, at 4.
25. Matthew Beard, Armed Police to Man Checkpoints in London as Drug‐Related
Crime Soars, INDEPENDENT (London), Sept. 7, 2002, at 2.
26 See Simon Heffer, If the State Fails Us, We Must Defend Ourselves, TELEGRAPH
ONLINE (London), Feb. 24, 2002, http://www.telegraph.co.uk/opinion/main.jhtml?xml=/
opinion/2002/02/24/do2401.xml; see also Ian Bell, Dublane Made Us All Think About
Gun Control . . . So What Went Wrong?, SUNDAY HERALD (Scotland), Feb. 24, 2007,
http://www.sundayherald.com/oped/opinion/display.var.1217778.0.dunblane_made
_us_all_think_about_gun_control_so_what_went_wrong.php; Comment, The Night
My Daughter Was Stabbed—And My Liberal Instincts Died, DAILY MAIL (London),
Mar. 5, 2007, http://www.dailymail.co.uk/pages/live/femail/article.html?in_ar‐ticle_
id=440318&in_page_id=1766&ito=1490.
27. See Melissa Kite, Tories Launch Bill to Give Householders the Power to Tackle In‐
truders, SUNDAY TELEGRAPH (London), Dec. 26, 2004, at 4; see also Renee Lerner, The
Worldwide Popular Revolt Against Proportionality in Self‐Defense Law, 2 J.L. ECON. &
POLʹY (2007).
1980s and 1990s. During these two decades, while Britain and

the Commonwealth were making lawful firearm ownership
increasingly difficult, more than 25 states in the United States
passed laws allowing responsible citizens to carry concealed
handguns. There are now 40 states where qualified citizens can
obtain such a handgun permit.28 As a result, the number of U.S.
citizens allowed to carry concealed handguns in shopping
malls, on the street, and in their cars has grown to 3.5 million
men and women.29 Economists John Lott and David Mustard
have suggested that these new laws contributed to the drop in
homicide and violent crime rates. Based on 25 years of corre‐
lated statistics from all of the more than 3,000 American coun‐
ties, Lott and Mustard conclude that adoption of these statutes
has deterred criminals from confrontation crime and caused
murder and violent crime to fall faster in states that adopted
this policy than in states that did not.30

28. In March 2006, Kansas and Nebraska became the 39th and 40th states, respec‐
tively, to pass “shall issue” concealed carry legislation. In Kansas, the state legisla‐
ture voted to overturn the governor’s veto of the bipartisan legislation. Kansas House
Overrides Concealed‐Guns‐Bill Veto, DESERET MORNING NEWS, Mar. 24, 2006. In Ne‐
braska, the governor signed the bill as passed by the state legislature. Kevin
O’Hanlon, Concealed‐Weapons Bill Adopted, LINCOLN JOURNAL STAR, Mar. 31, 2006.
29. Don Kates, The Limited Importance of Gun Control from a Criminological Perspec‐
tive, in SUING THE GUN INDUSTRY: A BATTLE AT THE CROSSROADS OF GUN CONTROL
AND MASS TORTS 62, 64 (Timothy D. Lytton ed., 2005).
30. See John R. Lott Jr. & David B. Mustard, Crime, Deterrence, and Right‐to‐Carry
Concealed Handguns, 26 J. LEGAL STUD. 1, 1 (1997); see also JOHN R. LOTT, JR., MORE
GUNS, LESS CRIME 19 (2d. ed. 2000). This conclusion is vehemently rejected by anti‐
gun advocates and academics who oppose armed self‐defense. See, e.g., Albert W.
Alschuler, Two Guns, Four Guns, Six Guns, More Guns: Does Arming the Public Reduce
Crime?, 31 VAL. U. L. REV. 365, 366 (1997); Ian Ayres & John J. Donohue III, Shooting
Down the ‘More Guns, Less Crime’ Hypothesis, 55 STAN. L. REV. 1193, 1197 (2003); Dan
A. Black & Daniel S. Nagin, Do Right‐to‐Carry Laws Deter Violent Crime?, 27 J. LEGAL
STUD. 209, 209 (1998); Franklin Zimring & Gordon Hawkins, Concealed Handguns:
The Counterfeit Deterrent, RESPONSIVE COMMUNITY, Spring 1997, at 46; Daniel W.
Webster, The Claims That Right‐to‐Carry Laws Reduce Violent Crime Are Unsubstanti‐
ated (Johns Hopkins Center for Gun Policy and Research, 1997).
Several critics have now replicated Lott’s work using additional or different data,
additional control variables, or new or different statistical techniques they deem
superior to those Lott used. Interestingly, the replications all confirm Lott’s general
conclusions; some even find that Lott underestimated the crime‐reductive effects of
allowing good citizens to carry concealed guns. See Jeffrey A. Miron, Violence, Guns,
and Drugs: A Cross‐Country Analysis, 44 J.L. & ECON. 615 (2001); David B. Mustard,
The Impact of Gun Laws on Police Deaths, 44 J.L. & ECON. 635 (2001); John R. Lott, Jr. &
John E. Whitley, Safe‐Storage Gun Laws: Accidental Deaths, Suicides, and Crime, 44 J.L.
& ECON. 659 (2001); Thomas B. Marvell, The Impact of Banning Juvenile Gun Posses‐
sion, 44 J.L. & ECON. 691 (2001); Jeffrey S. Parker, Guns, Crime, and Academics: Some
Reflections on the Gun Control Debate, 44 J.L. & ECON. 715 (2001); Bruce L. Benson &

As indicated in the preceding footnote, the notion that more
guns reduce crime is highly controversial. What the contro‐
versy has obscured from view is the corrosive effect of the Lott
and Mustard work on the tenet that more guns equal more
murder. As previously stated, adoption of state laws permit‐
ting millions of qualified citizens to carry guns has not resulted
in more murder or violent crime in these states. Rather, adop‐
tion of these statutes has been followed by very significant re‐
ductions in murder and violence in these states.
To determine whether this expansion of gun availability
caused reductions in violent crime requires taking account of
various other factors that might also have contributed to the
decline. For instance, two of Lottʹs major critics, Donohue and
Levitt, attribute much of the drop in violent crime that started
in 1990s to the legalization of abortion in the 1970s, which they
argue resulted in the non‐birth of vast numbers of children
who would have been disproportionately involved in violent
crime had they existed in the 1990s.31
The Lott‐Mustard studies did not address the Donohue‐
Levitt thesis. Lott and Mustard did account, however, for two
peculiarly American phenomena which many people believed
may have been responsible for the 1990s crime reduction: the
dramatic increase of the United States prison population and
the number of executions. The prison population in the United
States tripled during this time period, jumping from approxi‐
mately 100 prisoners per 100,000 in the late 1970s to more than
300 per 100,000 people in the general population in the early
1990s.32 In addition, executions in the United States soared

Brent D. Mast, Privately Produced General Deterrence, 44 J.L. & ECON. 725 (2001);
David E. Olson & Michael D. Maltz, Right‐to‐Carry Concealed Weapon Laws and Homi‐
cide in Large U.S. Counties: The Effect on Weapon Types, Victim Characteristics, and Vic‐
tim‐Offender Relationships, 44 J.L. & ECON. 747 (2001); Florenz Plassmann & T.
Nicolaus Tideman, Does the Right to Carry Concealed Handguns Deter Countable
Crimes? Only a Count Analysis Can Say, 44 J.L. & ECON. 771 (2001); Carlisle E. Moody,
Testing for the Effects of Concealed Weapons Laws: Specification Errors and Robustness, 44
J.L. & ECON. 799 (2001); see also Florenz Plassman & John Whitley, Confirming ‘More
Guns, Less Crime,’ 55 STAN. L. REV. 1313, 1316 (2003). In 2003, Lott reiterated and
extended his findings, which were subsequently endorsed by three Nobel laureates.
See JOHN R. LOTT, JR., THE BIAS AGAINST GUNS (2003).
31. See John J. Donohue III & Steven D. Levitt, The Impact of Legalized Abortion on
Crime, 116 Q. J. ECON. 379 (2001).
32. See Bureau of Justice Statistics, Key Facts at a Glance: Incarceration Rate, 1980–
2004 (Oct. 23, 2005), http://www.ojp.usdoj.gov/bjs/glance/tables/incrttab.htm, citing
ALLEN BECK & PAIGE HARRISON, BUREAU OF JUSTICE STATISTICS, CORRECTIONAL
POPULATIONS IN THE UNITED STATES 1997 (2000), available at

from approximately 5 per year in the early 1980s to more than
27 per year in the early 1990s.33 Neither of these trends is re‐
flected in Commonwealth countries.
Although the reason is thus obscured, the undeniable result
is that violent crime, and homicide in particular, has plum‐
meted in the United States over the past 15 years.34 The fall in
the American crime rate is even more impressive when com‐
pared with the rest of the world. In 18 of the 25 countries sur‐
veyed by the British Home Office, violent crime increased
during the 1990s.35 This contrast should induce thoughtful
people to wonder what happened in those nations, and to
question policies based on the notion that introducing increas‐
ingly more restrictive firearm ownership laws reduces violent
crime. Perhaps the United States is doing something right in
promoting firearms for law‐abiding responsible adults. Or per‐
haps the United States’ success in lowering its violent crime
rate relates to increasing its prison population or its death sen‐
tences.36 Further research is required to identify more precisely
which elements of the United States’ approach are the most
important, or whether all three elements acting in concert were
necessary to reduce violent crimes.
I. VIOLENCE: THE DECISIVENESS OF SOCIAL FACTORS
One reason the extent of gun ownership in a society does not
spur the murder rate is that murderers are not spread evenly
throughout the population. Analysis of perpetrator studies
shows that violent criminals—especially murderers—“almost

http://www.ojp.usdoj.gov/bjs/pub/pdf/cpus97.pdf, and ALLEN BECK & PAIGE
HARRISON, BUREAU OF JUSTICE STATISTICS, PRISONERS IN 2004 (2005), available at
http://www.ojp.usdoj.gov/bjs/pub/pdf/p04.pdf.).
33. THOMAS BONCZAR & TRACY L. SNELL, BUREAU OF JUSTICE STATISTICS
BULLETIN, CAPITAL PUNISHMENT 2003, (2004), available at http://www.ojp.usdoj.gov/
bjs/pub/pdf/cp03.pdf.
34. See generally FBI, VIOLENT CRIME, http://www.fbi.gov/ucr/05cius/offenses/
violent_crime/index.html; FBI, CRIME IN THE UNITED STATES BY VOLUME AND RATE PER
100,000 INHABITANTS, 1986–2005, http://www.fbi.gov/ucr/05cius/data/table_01.html.
35. See Gordon Barclay et al., International Comparisons of Criminal Justice Statistics
1999, HOME OFFICE STAT. BULL. (Research Development and Statistics, U.K. Home
Office, London, U.K.), 2001, available at http://www.homeoffice.gov.uk/rds/pdfs/
hosb601.pdf.
36. Several recent studies by economists calculate that each execution deters the
commission of 19 murders. See Cass R. Sunstein & Adrian Vermuele, Is Capital Pun‐
ishment Morally Required? Acts, Omissions, and Life‐Life Tradeoffs, 58 STAN. L. REV. 703
(2005).
uniformly have a long history of involvement in criminal behav‐
ior.”37 So it would not appreciably raise violence if all law‐
abiding, responsible people had firearms because they are not
the ones who rape, rob, or murder.38 By the same token, violent
crime would not fall if guns were totally banned to civilians. As
the respective examples of Luxembourg and Russia suggest,39
individuals who commit violent crimes will either find guns
despite severe controls or will find other weapons to use. 40
Startling as the foregoing may seem, it represents the cross‐
national norm, not some bizarre departure from it. If the man‐
tra “more guns equal more death and fewer guns equal less
death” were true, broad based cross‐national comparisons
should show that nations with higher gun ownership per cap‐
ita consistently have more death. Nations with higher gun
ownership rates, however, do not have higher murder or sui‐
cide rates than those with lower gun ownership. Indeed many
high gun ownership nations have much lower murder rates.
Consider, for example, the wide divergence in murder rates
among Continental European nations with widely divergent
gun ownership rates.
The non‐correlation between gun ownership and murder
is reinforced by examination of statistics from larger num‐
bers of nations across the developed world. Comparison of
“homicide and suicide mortality data for thirty‐six nations
(including the United States) for the period 1990–1995” to
gun ownership levels showed “no significant (at the 5%
level) association between gun ownership levels and the to‐
tal homicide rate.”41 Consistent with this is a later European
study of data from 21 nations in which “no significant corre‐
lations [of gun ownership levels] with total suicide or homi‐
cide rates were found.”42

37. See Delbert S. Elliott, Life‐Threatening Violence is Primarily a Crime Problem: A
Focus on Prevention, 69 COLO. L. REV. 1081, 1089 (1998) (emphasis added).
38. See infra Part III.
39. See supra notes 3–9 and Table 1.
40. See supra Table 1 and infra Tables 2–3.
41. KLECK, supra note 8, at 254. The study also found no correlation to suicide
rates. Id.
42. Martin Killias et al., Guns, Violent Crime, and Suicide in 21 Countries, 43 CAN. J.
CRIMINOLOGY & CRIM. JUST. 429, 430 (2001). It bears emphasis that the authors, who
are deeply anti‐gun, emphasize the “very strong correlations between the presence
of guns in the home and suicide committed with a gun”—as if there were some import

II. ASKING THE WRONG QUESTION
However unintentionally, the irrelevance of focusing on

weaponry is highlighted by the most common theme in the
more guns equal more death argument. Epitomizing this
theme is a World Health Organization (WHO) report assert‐
ing, “The easy availability of firearms has been associated
with higher firearm mortality rates.”43 The authors, in noting
that the presence of a gun in a home corresponds to a higher
risk of suicide, apparently assume that if denied firearms,
potential suicides will decide to live rather than turning to
the numerous alternative suicide mechanisms. The evidence,
however, indicates that denying one particular means to
people who are motivated to commit suicide by social, eco‐
nomic, cultural, or other circumstances simply pushes them
to some other means.44 Thus, it is not just the murder rate in
gun‐less Russia that is four times higher than the American
rate; the Russian suicide rate is also about four times higher
than the American rate.45

to the death being by gun rather than by hanging, poison, or some other means. Id.;
see also infra Part III.
43. WORLD HEALTH ORGANIZATION, SMALL ARMS AND GLOBAL HEALTH 11 (2001)
(emphasis added). This irrelevancy is endlessly repeated. See, e.g., Wendy Cukier,
Small Arms and Light Weapons: A Public Health Approach, 9 BROWN J. WORLD AFF. 261,
266, 267 (2002) (“Research has shown that rates of small arms death and injury are
linked to small arms accessibility . . . . In industrialized countries, studies have
shown that accessibility is related to firearm death rates . . . . Other approaches have
examined the rates of death from firearms across regions, cities, high income coun‐
tries, and respondents to victimization surveys.” (emphasis added) (internal cita‐
tions omitted); see also Neil Arya, Confronting the Small Arms Pandemic 324 BRITISH
MED. J. 990 (2002); E.G. Krug et al., Firearm‐Related Deaths in the United States and 35
Other High and Upper‐Middle‐Income Countries, 27 INT’L J. EPIDEMIOLOGY 214 (1988).
44. See JACOBS, supra note 11, at 120 (“[I]f the Brady Law did have the effect of
modestly reducing firearms suicides . . . this effect was completely offset by an in‐
crease of the same magnitude in nonfirearm suicide” resulting in the same number
of deaths); see also KLECK, supra note 8, at 265–92 (summarizing and reviewing stud‐
ies regarding guns and suicide). Indeed, though without noting the significance, the
WHO report states that out of sample of 52 countries, “firearms accounted for only
one‐fifth of all suicides, just ahead of poisoning . . . . [Self‐] strangulation, [i.e. hang‐
ing] was the most frequently used method of suicide.” WORLD HEALTH
ORGANIZATION, supra note 43, at 3.
45. In 1999, the latest year for which we have Russian data, the American suicide
rate was 10.7 per 100,000 people, while the Russian suicide rate was almost 41 per
100,000 people. William Alex Pridemore & Andrew L. Spivak, Patterns of Suicide
Mortality in Russia, 33 SUICIDE & LIFE‐THREATENING BEHAVIOR 132, 133 (2003);
Donna L. Hoyert et al., Deaths: Final Data for 1999, NAT’L VITAL STAT. REP., Sept. 21,
2001, at 6.
There is no social benefit in decreasing the availability of
guns if the result is only to increase the use of other means of
suicide and murder, resulting in more or less the same
amount of death. Elementary as this point is, proponents of
the more guns equal more death mantra seem oblivious to it.
One study asserts that Americans are more likely to be shot
to death than people in the world’s other 35 wealthier na‐
tions.46 While this is literally true, it is irrelevant—except,
perhaps to people terrified not of death per se but just death
by gunshot. A fact that should be of greater concern—but
which the study fails to mention—is that per capita murder
overall is only half as frequent in the United States as in sev‐
eral other nations where gun murder is rarer, but murder by
strangling, stabbing, or beating is much more frequent.47
Of course, it may be speculated that murder rates around
the world would be higher if guns were more available. But
there is simply no evidence to support this. Like any specu‐
lation, it is not subject to conclusive disproof; but the Euro‐
pean data in Table 1 and the studies across 36 and 21 nations
already discussed show no correlation of high gun owner‐
ship nations and greater murder per capita or lower gun
ownership nations and less murder per capita.48
To reiterate, the determinants of murder and suicide are
basic social, economic, and cultural factors, not the preva‐
lence of some form of deadly mechanism. In this connection,
recall that the American jurisdictions which have the highest
violent crime rates are precisely those with the most strin‐
gent gun controls.49 This correlation does not necessarily

46. See Krug et al., supra note 42, at 218–19.
47. Id. at 216. Two of those nations, Brazil and Estonia, had more than twice the
overall murder rates of the United States. David C. Stolinsky, America: The Most
Violent Nation?, 5 MED. SENTINEL 199, 200 (2000). Readers may question the value of
comparing the United States to those particular nations; however, this comparison
was first suggested by Krug. Krug et al., supra note 43, at 215 (using thirty‐six coun‐
tries, having among the highest GNP per capita as listed in the World Bankʹs 1994
World Development Report). All we have done is provide full murder rate informa‐
tion for these comparisons.
48. KLECK, supra note 8, at 254; Killias et al., supra note 41, at 430.
49. See infra notes 128‐30 and accompanying text. For at least thirty years, gun ad‐
vocates have echoed in more or less identical terms the observation that twenty
percent of American homicide is concentrated in four cities with the nationʹs most
restrictive gun laws. See Firearms Legislation: Hearing Before the Subcomm. on Crime of
the H. Comm. on the Judiciary, 94th Cong. 2394 (1975) (statement of Neal Knox). In
October 2000, the head of a gun advocacy group ridiculed a Handgun Control

prove gun advocates’ assertion that gun controls actually

encourage crime by depriving victims of the means of self‐
defense. The explanation of this correlation may be political
rather than criminological: jurisdictions afflicted with violent
crime tend to severely restrict gun ownership. This, how‐
ever, does not suppress the crime, for banning guns cannot
alleviate the socio‐cultural and economic factors that are the
real determinants of violence and crime rates. 50
Table 2: Murder Rates of European Nations that Ban
Handguns as Compared to Their Neighbors that Allow Handguns
(rates are per 100,000 persons)
Nation Handgun Policy Murder Rate Year

A. Belarus banned 10.40 late 1990s
[Neighboring countries with gun law and murder rate data available]
Poland allowed 1.98 2003
Russia banned 20.54 2002
B. Luxembourg banned 9.01 2002
Belgium allowed 1.70 late 1990s
France allowed 1.65 2003
Germany allowed 0.93 2003
C. Russia banned 20.54 2002
Finland allowed 1.98 2004
Norway allowed 0.81 2001
Notes: This table covers all the European nations for which the infor‐
mation given is available. As in Table 1, the homicide rate data comes
from an annually published report, CANADIAN CENTRE FOR JUSTICE
STATISTICS, HOMICIDE IN CANADA, JURISTAT.

Once again, we are not arguing that the data in Table 2
shows that gun control causes nations to have much higher

“scorecard” for its misleading attempts to inversely correlate violent crime rates to
the extent of the various states’ gun controls. He points out that, in fact, the states
with the most restrictive gun laws consistently have higher murder rates than states
with less restrictive laws, while those with the least controls had the lowest homi‐
cide rates. Larry Pratt, HCI Scorecard (2000), http://gunowners.org/op0042.htm; see
also infra note 131.
50. It is noteworthy that the correlation between more gun control and more crime
seems to hold true in other nations, though much less strikingly than in the United
States. See Miron, supra note 30, at 628.
murder rates than neighboring nations that permit handgun

ownership. Rather, we assert a political causation for the ob‐
served correlation that nations with stringent gun controls tend
to have much higher murder rates than nations that allow
guns. The political causation is that nations which have vio‐
lence problems tend to adopt severe gun controls, but these do
not reduce violence, which is determined by basic socio‐
cultural and economic factors.
The point is exemplified by the conclusions of the premier
study of English gun control. Done by a senior English police
official as his thesis at the Cambridge University Institute of
Criminology and later published as a book, it found (as of the
early 1970s), “Half a century of strict controls . . . has ended,
perversely, with a far greater use of [handguns] in crime than
ever before.” 51 The study also states that:
No matter how one approaches the figures, one is forced to
the rather startling conclusion that the use of firearms in
crime was very much less [in England before 1920] when
there were no controls of any sort and when anyone, con‐
victed criminal or lunatic, could buy any type of firearm
without restriction.52
Of course the point of this analysis is not that the law should
allow lunatics and criminals to own guns. The point is that vio‐
lence will be rare when the basic socio‐cultural and economic
determinants so dictate; and conversely, crime will rise in re‐
sponse to changes in those determinants—without much re‐
gard to the mere availability of some particular weaponry or
the severity of laws against it.
III. DO ORDINARY PEOPLE MURDER?
The “more guns equal more death” mantra seems plausible

only when viewed through the rubric that murders mostly in‐
volve ordinary people who kill because they have access to a
firearm when they get angry. If this were true, murder might
well increase where people have ready access to firearms, but
the available data provides no such correlation. Nations and

51. COLIN GREENWOOD, FIREARMS CONTROL: A STUDY OF ARMED CRIME AND
FIREARMS CONTROL IN ENGLAND AND WALES 243 (1972).
52. Id.
areas with more guns per capita do not have higher murder
rates than those with fewer guns per capita.53
Nevertheless, critics of gun ownership often argue that a
“gun in the closet to 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;”54 that banning handgun posses‐
sion only for those with criminal records will “fail to protect us
from the most likely source of handgun murder: ordinary citi‐
zens;”55 that “most gun‐related homicides . . . are the result of
impulsive actions taken by individuals who have little or no
criminal background or who are known to the victims;”56 that
“the majority of firearm homicide[s occur] . . . not as the result
of criminal activity, but because of arguments between people
who know each other;”57 that each year there are thousands of
gun murders “by law‐abiding citizens who might have stayed
law‐abiding if they had not possessed firearms.”58
These comments appear to rest on no evidence and actually con‐
tradict facts that have so uniformly been established by homicide
studies dating back to the 1890s that they have become “crimino‐
logical axioms.”59 Insofar as studies focus on perpetrators, they
show that neither a majority, nor many, nor virtually any murder‐
ers are ordinary “law‐abiding citizens.”60 Rather, almost all mur‐
derers are extremely aberrant individuals with life histories of
violence, psychopathology, substance abuse, and other dangerous
behaviors. “The vast majority of persons involved in life‐
threatening violence have a long criminal record with many prior
contacts with the justice system.”61 “Thus homicide—[whether] of a

53. See supra Tables 1–2 and notes 10–15; see infra Table 3 and notes 125–127.
54. David Kairys, A Carnage in the Name of Freedom, PHILADELPHIA INQUIRER, Sept.
12, 1988, at A15 (emphasis added), quoted in Frank J. Vandall, A Preliminary Consid‐
eration of Issues Raised in the Firearms Sellers Immunity Bill, 38 AKRON L. REV. 113, 118
n.28 (2005).
55. Nicholas Dixon, Why We Should Ban Handguns in the United States, 12 ST. LOUIS
U. PUB. L. REV. 243, 265–66 (1993) (emphasis added), quoted in Vandall, supra note 54,
at 119, n.32.
56. ROBERT J. SPITZER, THE POLITICS OF GUN CONTROL 147 (3rd ed. 1995) (empha‐
sis added).
57. Violence Policy Center, Who Dies? A Look at Firearms Death and Injury in
America, http://www.vpc.org/studies/whointro.htm (last visited Nov. 17, 2006).
58. Natl Coalition to Ban Handguns, supra note 21 (emphasis added).
59. See David M. Kennedy & Anthony J. Braga, Homicide in Minneapolis: Research
for Problem Solving, 2 HOMICIDE STUD. 263, 267 (1998).
60. See Elliott, supra note 37, at 1093.
61. Id.
stranger or [of] someone known to the offender—‘is usually part of
a pattern of violence, engaged in by people who are known . . . as
violence prone.’”62 Though only 15% of Americans over the age of
15 have arrest records,63 approximately 90 percent of “adult mur‐
derers have adult records, with an average adult criminal career
[involving crimes committed as an adult rather than a child] of six
or more years, including four major adult felony arrests.”64 These
national statistics dovetail with data from local nineteenth and
twentieth century studies. For example: victims as well as offenders
[in 1950s and 1960s Philadelphia murders] . . . tended to be people
with prior police records, usually for violent crimes such as as‐
sault.”65 “The great majority of both perpetrators and victims of
[1970s Harlem] assaults and murders had previous [adult] arrests,
probably over 80% or more.”66 Boston police and probation officers
in the 1990s agreed that of those juvenile‐perpetrated murders
where all the facts were known, virtually all were committed by
gang members, though the killing was not necessarily gang‐
directed. 67 One example would be a gang member who stabs his
girlfriend to death in a fit of anger.68 Regardless of their arrests for
other crimes, 80% of 1997 Atlanta murder arrestees had at least one
earlier drug offense with 70% having 3 or more prior drug of‐
fenses.69 A New York Times study of the 1,662 murders committed in
that city in the years 2003–2005 found that “[m]ore than 90 percent
of the killers had criminal records.”70 Baltimore police figures show
that “92 percent of murder suspects had [prior] criminal records in
2006.”71 Several of the more recent homicide studies just reviewed

62. GERALD D. ROBIN, VIOLENT CRIME AND GUN CONTROL 48 (1991) (quoting Gary
Kleck, The Assumptions of Gun Control, in FIREARMS AND VIOLENCE 23, 43 (Don B.
Kates ed., 1984)).
63. Mark Cooney, The Decline of Elite Homicide, 35 CRIMINOLOGY 381, 386 (1997).
64. GARY KLECK & DON B. KATES, ARMED: NEW PERSPECTIVES ON GUN CONTROL
20 (2001).
65. ROGER LANE, MURDER IN AMERICA: A HISTORY 259 (1997).
66. A. SWERSKEY & E. ENLOE, HOMICIDE IN HARLEM 17 (1975).
67. Anthony A. Braga et al., Youth Homicide in Boston: An Assessment of Supplemen‐
tary Homicide Report Data, 3 HOMICIDE STUDIES 277, 283–84 (1999).
68. See id.
69. Dean G. Rojek, The Homicide and Drug Connection, in THE VARIETIES OF
HOMICIDE AND ITS RESEARCH: PROCEEDINGS OF THE 1999 MEETING OF THE HOMICIDE
RESEARCH WORKING GROUP 128 (P.H. Blackman et al. eds., 2000) [hereinafter THE
VARIETIES OF HOMICIDE].
70. Jo Craven McGinty, New York Killers, and Those Killed, by the Numbers, N.Y.
TIMES, April 28, 2006, at A1.
71. Gus G. Sentementes, Patterns Persist in City Killings, BALTIMORE SUN, Jan. 1,
2007, at A1.
were done at the Kennedy School of Government at Harvard and
found almost all arrested murderers to have earlier arrests.72
That murderers are not ordinary, law‐abiding responsible
adults is further documented in other sources. Psychological
studies of juvenile murderers variously find that at least 80%, if
not all, are psychotic or have psychotic symptoms.73 Of Massa‐
chusetts domestic murderers in the years 1991–1995, 73.7% had
a “prior [adult] criminal history,” 16.5% had an active restrain‐
ing order registered against them at the time of the homicide,
and 46.3% of the violent perpetrators had had a restraining or‐
der taken out against them sometime before their crime.74
This last study is one of many exposing the false argument
that a significant number of murders involve ordinary people
killing spouses in a moment of rage. Although there are many
domestic homicides, such murders do not occur frequently in
ordinary families, nor are the murderers ordinary, law‐abiding
adults. “The day‐to‐day reality is that most family murders are
prefaced by a long history of assaults.”75 One study of such
murders found that “a history of domestic violence was pre‐
sent in 95.8%” of cases.76 These findings are a routine feature of
domestic homicide studies: “[domestic] partner homicide is
most often the final outcome of chronic women battering”;77
based on a study from Kansas City, 90% of all the family homi‐

72. Anthony A. Braga et al., Understanding and Preventing Gang Violence: Problem
Analysis and Response Development in Lowell, Massachusetts, 9 POLICE Q. 20, 29–31
(2006) (“Some 95% of homicide offenders, 82% of aggravated assault offenders, 65%
of homicide victims, and 45% of aggravated gun assault victims were arraigned at
least once in Massachusetts courts before they committed their crime or were vic‐
timized. Individuals that were previously known to the criminal justice system were
involved in a wide variety of offenses and, on average, committed many prior
crimes . . . . On average, aggravated gun assault offenders had been arraigned for 12
prior offenses, homicide offenders had been arraigned for 9 prior offenses . . . .”).
73. Wade C. Myers & Kerrilyn Scott, Psychotic and Conduct Disorder Symptoms in
Juvenile Murderers, 2 HOMICIDE STUD. 160, 161–62 (1998).
74. Linda Langford et al., Criminal and Restraining Order Histories of Intimate Part‐
ner‐Related Homicide Offenders in Massachusetts, 1991‐1995, in THE VARIETIES OF
HOMICIDE, supra note 69, at 51, 55, 59.
75. Murray A. Straus, Domestic Violence and Homicide Antecedents, 62 BULL. N.Y.
ACAD. MED. 446, 454 (1986); see also Murray A. Straus, Medical Care Costs of Intrafam‐
ily Assault and Homicide, 62 BULL. N.Y. ACAD. MED. 556, 557 (1986).
76. Paige Hall Smith et al., Partner Homicide in Context, 2 HOMICIDE STUD. 400, 410
(1998) (reporting cases only where there was sufficient background information on
the parties).
77. Id. at 411.
cides were preceded by previous disturbances at the same ad‐
dress, with a median of 5 calls per address.”78
The only kind of evidence cited to support the myth that
most murderers are ordinary people is that many murders
arise from arguments or occur in homes and between acquaint‐
ances.79 These bare facts are only relevant if one assumes that
criminals do not have acquaintances or homes or arguments.
Of the many studies belying this, the broadest analyzed a
year’s national data on gun murders occurring in homes and
between acquaintances. It found “the most common victim‐
offender relationship” was “where both parties . . . knew one
another because of prior illegal transactions.”80
Thus the term “acquaintance homicide” does not refer solely
to murders between ordinary acquaintances. Rather it encom‐
passes, for example: drug dealers killed by competitors or cus‐
tomers, gang members killed by members of the same or rival
gangs, and women killed by stalkers or abusers who have bru‐
talized them on earlier occasions, all individuals for whom fed‐
eral and state laws already prohibit gun possession.81
Obviously there are certain people who should not be al‐
lowed to own any deadly instrument. Reasonable as such pro‐
hibitions are, it is unrealistic to think those people will comply
with such restrictions any more readily than they do with laws
against violent crime.82 In any event, studies analyzing ac‐

78. ROBIN, supra note 62, at 47–48; see also Kathryn E. Moracco et al., Femicide in
North Carolina, 1991‐1993, 2 HOMICIDE STUD. 422, 441 (1998).
79. See, e.g., SPITZER, supra note 56; Jeremiah A. Barondess, Letter to the Editor,
Firearm Violence and Public Health, 272 J. AM. MED. ASS’N 1406, 1409 (1994) (respond‐
ing to criticism of his article, Karl P. Adler et al., Firearm Violence and Public Health:
Limiting the Availability of Guns, 271 J. AM. MED. ASS’N 1281 (1994)); Kairys, supra
note 54.
80. KLECK, supra note 8, at 236 (analyzing the U.S. Bureau of Justice Statistics data
on murder defendants being prosecuted in 33 U.S. urban counties).
81. Current federal law prohibits gun possession by minors, drug addicts, and
persons who have been involuntarily committed to mental institutions or convicted
of felonies or domestic violence misdemeanors. 18 U.S.C. § 922(g) (2000). As to state
gun laws, see, for example, CAL. PENAL CODE §§ 12021, 12072, 12101, 12551 (Deer‐
ing 2006). For a summary of the general patterns of federal and state gun laws, see
JACOBS, supra note 11, at 19–35.
82. See WRIGHT ET AL., supra note 11, at 137–38 (“[T]here is no good reason to sup‐
pose that people intent on arming themselves for criminal purposes would not be
able to do so even if the general availability of firearms to the larger population were
sharply restricted. Here it may be appropriate to recall the First Law of Economics, a
law whose operation has been sharply in evidence in the case of Prohibition, mari‐
juana and other drugs, prostitution, pornography, and a host of other banned arti‐

quaintance homicide suggest there is no reason for laws pro‐
hibiting gun possession by ordinary, law‐abiding responsible
adults because such people virtually never murder. If one ac‐
cepts that such adults are far more likely to be victims of vio‐
lent crime than to commit it, disarming them becomes not just
unproductive but counter‐productive.83
IV. MORE GUNS, LESS CRIME?
Anti‐gun activists are not alone in their belief that widespread
firearm ownership substantially affects violent crime rates. The
same understanding also characterizes many pro‐gun activists. Of
course, pro‐gun activists’ belief leads them to the opposite conclu‐
sion: that widespread firearm ownership reduces violence by de‐
terring criminals from confrontation crimes and making more
attractive such nonconfrontation crimes as theft from unoccupied
commercial or residential premises. Superficially, the evidence for
this belief seems persuasive. Table 1, for instance, shows that Den‐
mark has roughly half the gun ownership rate of Norway, but a
50% higher murder rate, while Russia has only one‐ninth Norway’s

cles and substances—namely, that demand creates its own supply. There is no evi‐
dence anywhere to show that reducing the availability of firearms in general likewise
reduces their availability to persons with criminal intent, or that persons with crimi‐
nal intent would not be able to arm themselves under any set of general restrictions
on firearms.”).
83. This Article will not discuss the defensive use of firearms beyond making the
following observations: while there is a great deal of controversy about the subject,
it is a misleading controversy in which anti‐gun advocatesʹ deep ethical or moral
objections to civilian self‐defense are presented in the guise of empirical argument.
The empirical evidence unquestionably establishes that gun ownership by prospec‐
tive victims not only allows them to resist criminal attack, but also deters violent
criminals from attacking them in the first place. See JOSEPH F. SHELEY & JAMES D.
WRIGHT, IN THE LINE OF FIRE: YOUTHS, GUNS, AND VIOLENCE IN URBAN AMERICA 63
(1995), and JAMES D. WRIGHT & PETER H. ROSSI, ARMED AND CONSIDERED
DANGEROUS: A SURVEY OF FELONS AND THEIR FIREARMS 154 (1986) for a discussion
of Dept. of Justice‐funded surveys of incarcerated adult and juvenile felons. See also
LOTT, THE BIAS AGAINST GUNS, supra note 30, at 8–11, 227–40; David B. Kopel, Law‐
yers, Guns, and Burglars, 43 ARIZ. L. REV. 345 (2001); Lawrence Southwick, Jr., Self‐
Defense with Guns: The Consequences, 28 J. CRIM. JUST. 351 (2000).
The legitimate question is not whether victim gun possession allows for self‐
defense and deters criminal violence, but how extensive and important these bene‐
fits are. See KLECK & KATES, supra note 64, at 213–342; LOTT, supra note 11; Philip J.
Cook & Jens Ludwig, Defensive Gun Uses: New Evidence from a National Survey, 14 J.
QUANTITATIVE CRIMINOLOGY 111 (1998); Philip J. Cook & Jens Ludwig, Guns in
America: National Survey on Private Ownership and Use of Firearms, Nat’l Inst. Just.:
Research in Brief (U.S. Dep’t of Justice, Washington, D.C., 1997); Marvin E. Wolf‐
gang, A Tribute to a View I Have Opposed, 86 J. CRIM. L. & CRIMINOLOGY 188 (1995).
gun ownership rate but a murder rate 2500% higher. Looking at
Tables 1–3, it is easy to find nations in which very high gun owner‐
ship rates correlate with very low murder rates, while other nations
with very low gun ownership rates have much higher murder
rates. Moreover, there is not insubstantial evidence that in the
United States widespread gun availability has helped reduce mur‐
der and other violent crime rates. On closer analysis, however, this
evidence appears uniquely applicable to the United States.
More than 100 million handguns are owned in the United
States84 primarily for self‐defense,85 and 3.5 million people have
permits to carry concealed handguns for protection.86 Recent
analysis reveals “a great deal of self‐defensive use of firearms” in
the United States, “in fact, more defensive gun uses [by victims]
than crimes committed with firearms.”87 It is little wonder that the
National Institute of Justice surveys among prison inmates
find that large percentages report that their fear that a victim
might be armed deterred them from confrontation crimes.
“[T]he felons most frightened ‘about confronting an armed
victim’ were those from states with the greatest relative
number of privately owned firearms.” Conversely, robbery
is highest in states that most restrict gun ownership.88
Concomitantly, a series of studies by John Lott and his coauthor
David Mustard conclude that the issuance of millions of permits
to carry concealed handguns is associated with drastic declines in
American homicide rates.89
Ironically, to detail the American evidence for widespread de‐
fensive gun ownership’s deterrent value is also to raise questions
about how applicable that evidence would be even to the other
nations that have widespread gun ownership but low violence.
There are no data for foreign nations comparable to the American
data just discussed. Without such data, we cannot know whether
millions of Norwegians own handguns and carry them for protec‐

84. Kates, supra note 29, at 63.
85. KLECK, supra note 8, at 74 (collecting survey responses).
86. Kates, supra note 29, at 64.
87. JACOBS, supra note 11, at 14 (collecting studies).
88. Kates, supra note 29, at 70 (collecting studies).
89. LOTT, supra note 11; John R. Lott & David B. Mustard, Crime, Deterrence, and
Right‐to‐Carry, 26 J. LEGAL STUD. 1 (1997); David B. Mustard, Culture Affects Our
Beliefs About Firearms, But Data are Also Important, 151 U. PENN. L. REV. 1387 (2003).
These studies are highly controversial. See Kates, supra note 29, at 70–71, for discus‐
sion of critics and criticisms.
tion, thereby deterring Norwegian criminals from committing

violent crimes. Nor can we know whether guns are commonly
kept for defense in German homes and stores, thus preventing
German criminals from robbing them.
Moreover, if the deterrent effect of gun ownership accounts for
low violence rates in high gun ownership nations other than the
United States, one wonders why that deterrent effect would be
amplified there. Even with the drop in United States murder rates
that Lott and Mustard attribute to the massive increase in gun
carry licensing, the United States murder rate is still eight times
higher than Norway’s—even though the U.S. has an almost 300%
higher rate of gun ownership. That is consistent with the points
made above. Murder rates are determined by socio‐economic and
cultural factors. In the United States, those factors include that the
number of civilian‐owned guns nearly equals the population—
triple the ownership rate in even the highest European gun‐
ownership nations—and that vast numbers of guns are kept for
personal defense. That is not a factor in other nations with com‐
paratively high firearm ownership. High gun ownership may
well be a factor in the recent drastic decline in American homi‐
cide. But even so, American homicide is driven by socio‐economic
and cultural factors that keep it far higher than the comparable
rate of homicide in most European nations.
In sum, though many nations with widespread gun ownership
have much lower murder rates than nations that severely restrict
gun ownership, it would be simplistic to assume that at all times
and in all places widespread gun ownership depresses violence by
deterring many criminals into nonconfrontation crime. There is
evidence that it does so in the United States, where defensive gun
ownership is a substantial socio‐cultural phenomenon. But the
more plausible explanation for many nations having widespread
gun ownership with low violence is that these nations never had
high murder and violence rates and so never had occasion to enact
severe anti‐gun laws. On the other hand, in nations that have ex‐
perienced high and rising violent crime rates, the legislative reac‐
tion has generally been to enact increasingly severe antigun laws.
This is futile, for reducing gun ownership by the law‐abiding citi‐
zenry—the only ones who obey gun laws—does not reduce vio‐
lence or murder. The result is that high crime nations that ban guns
to reduce crime end up having both high crime and stringent gun
laws, while it appears that low crime nations that do not signifi‐
cantly restrict guns continue to have low violence rates.
Thus both sides of the gun prohibition debate are likely
wrong in viewing the availability of guns as a major factor in
the incidence of murder in any particular society. Though
many people may still cling to that belief, the historical, geo‐
graphic, and demographic evidence explored in this Article
provides a clear admonishment. Whether gun availability is
viewed as a cause or as a mere coincidence, the long term
macrocosmic evidence is that gun ownership spread widely
throughout societies consistently correlates with stable or
declining murder rates. Whether causative or not, the consis‐
tent international pattern is that more guns equal less mur‐
der and other violent crime. Even if one is inclined to think
that gun availability is an important factor, the available in‐
ternational data cannot be squared with the mantra that
more guns equal more death and fewer guns equal less
death. Rather, if firearms availability does matter, the data
consistently show that the way it matters is that more guns
equal less violent crime.
V. GEOGRAPHIC, HISTORICAL AND
DEMOGRAPHIC PATTERNS
If more guns equal more death and fewer guns equal less
death, it should follow, all things being equal, (1) that geo‐
graphic areas with higher gun ownership should have more
murder than those with less gun ownership; (2) that demo‐
graphic groups with higher gun ownership should be more
prone to murder than those with less ownership; and (3) that
historical eras in which gun ownership is widespread should
have more murder than those in which guns were fewer or
less widespread. As discussed earlier, these effects are not
present. Historical eras, demographic groups, and geo‐
graphic areas with more guns do not have more murders
than those with fewer guns. Indeed, those with more guns
often, or even generally, have fewer murders.
Of course, all other things may not be equal. Obviously,
many factors other than guns may promote or reduce the
number of murders in any given place or time or among par‐
ticular groups. And it may be impossible even to identify
these factors, much less to take account of them all. Thus any
conclusions drawn from the kinds of evidence presented ear‐
lier in this paper must necessarily be tentative.
Acknowledging this does not, however, blunt the force of
two crucial points. The first regards the burden of proof.
Those who assert the mantra, and urge that public policy be
based on it, bear the burden of proving that more guns do
equal more death and fewer guns equal less death. But they
cannot bear that burden because there simply is no large
number of cases in which the widespread prevalence of
guns among the general population has led to more mur‐
der. By the same token, but even more importantly, it can‐
not be shown consistently that a reduction in the number of
guns available to the general population has led to fewer
deaths. Nor is the burden borne by speculating that the rea‐
son such cases do not appear is that other factors always
intervene.
The second issue, allied to the burden of proof, regards
plausibility. On their face, the following facts from Tables 1
and 2 suggest that gun ownership is irrelevant, or has little
relevance, to murder: France and neighboring Germany
have exactly the same, comparatively high rate of gun own‐
ership, yet the French murder rate is nearly twice the Ger‐
man; France has infinitely more gun ownership than
Luxembourg, which nevertheless has a murder rate five
times greater, though handguns are illegal and other types
of guns sparse; Germany has almost double the gun owner‐
ship rate of neighboring Austria yet a similarly very low
murder rate; the Norwegian gun ownership rate is over
twice the Austrian rate, yet the murder rates are almost
identical.
And then there is Table 3, which shows Slovenia, with 66%
more gun ownership than Slovakia, nevertheless has roughly
one‐third less murder per capita; Hungary has more than 6
times the gun ownership rate of neighboring Romania but a
lower murder rate; the Czech Republic’s gun ownership rate
is more than 3 times that of neighboring Poland, but its mur‐
der rate is lower; Poland and neighboring Slovenia have ex‐
actly the same murder rate, though Slovenia has over triple
the gun ownership per capita.
Table 3: Eastern Europe Gun Ownership and Murder Rates
(rates given are per 100,000 people and in descending order)
Nation Murder Rate Rate of Gun Ownership

Russia 20.54* [2002] 4,000
Moldova 8.13** [2000] 1,000
Slovakia 2.65** [2000] 3,000
Romania 2.50** [2000] 300
Macedonia 2.31** [2000] 16,000
Hungary 2.22† [2003] 2,000
Finland 1.98‡ [2004] 39,000
Poland 1.79† [2003] 1,500
Slovenia 1.81** [2000] 5,000
Cz. Republic 1.69** [2000] 5,000
Greece 1.12† [2003] 11,000
Notes: This table covers all the Eastern European nations for which we have data
regarding both gun ownership and murder rates. Gun ownership data comes from
GRADUATE INSTITUTE OF INTERNATIONAL STUDIES, SMALL ARMS SURVEY (2003).
* CANADIAN CENTRE FOR JUSTICE STATISTICS, HOMICIDE IN CANADA, 2002,
JURISTAT at 3.
** United Nations Office on Drugs and Crime, The Seventh United Nations Survey
on Crime Trends and the Operations of Criminal Justice Systems (1998–2000), Mar. 31,
2004, at 82, 260, 287, 370, 405, 398.
†
CANADIAN CENTRE FOR JUSTICE STATISTICS, HOMICIDE IN CANADA, 2003,
JURISTAT at 3.
‡
CANADIAN CENTRE FOR JUSTICE STATISTICS, HOMICIDE IN CANADA, 2004,
JURISTAT at 3.

On their face, Tables 1, 2, and 3 and the comparisons gleaned
from them suggest that gun ownership is irrelevant, or has little
relevance, to murder. Historical and demographic comparisons
offer further evidence. Again, all the data may be misleading. It is
conceivable that more guns do equal more murder, but that this
causation does not appear because some unidentifiable extrane‐
ous factor always intervenes. That is conceivable, but ultimately
unlikely. As Hans Toch, a senior American criminologist who 35
years ago endorsed handgun prohibition and confiscation, but
then recanted based on later research, argues “it is hard to explain
that where firearms are most dense, violent crime rates are lowest
and where guns are least dense, violent crime rates are highest.”90

90. Toch & Lizotte, supra note 11, at 232. Professor Toch was a consultant to the
1960s Eisenhower Commission, and until the 1990s he endorsed its conclusions that

A. Demographic Patterns
Contrary to what should be the case if more guns equal more
death, there are no “consistent indications of a link between
gun ownership and criminal or violent behavior by owners;” in
fact, gun ownership is “higher among whites than among
blacks, higher among middle‐aged people than among young
people, higher among married than among unmarried people,
higher among richer people than poor”—all “patterns that are
the reverse of the way in which criminal behavior is distrib‐
uted.”91
These conclusions are reinforced by focusing on patterns of
African‐American homicide. Per capita, African‐American
murder rates are much higher than the murder rate for
whites.92 If more guns equal more death, and fewer guns equal
less, one might assume gun ownership is higher among Afri‐
can‐Americans than among whites, but in fact African‐
American gun ownership is markedly lower than white gun
ownership.93

widespread handgun ownership causes violence and that reducing ownership
would reduce violence. Franklin Zimring, one of the architects of those conclusions,
has admitted that they were made speculatively and essentially without an empiri‐
cal basis. FRANKLIN E. ZIMRING & GORDON HAWKINS, THE CITIZEN’S GUIDE TO GUN
CONTROL xi‐xii (1987) (“In the 1960s after the assassinations of President John F.
Kennedy, Dr. Martin Luther King, Jr., and Senator Robert F. Kennedy, it [gun con‐
trol] became a major subject of public passion and controversy . . . [sparking a de‐
bate that] has been heated, acrimonious and polarized . . . . It began in a factual
vacuum [in which] . . . neither side felt any great need for factual support to buttress
foregone conclusions. In the 1960s, there was literally no scholarship on the relationship
between guns and violence and the incidence or consequences of interpersonal violence, and
no work in progress.” (emphasis added)).
As for the findings of the subsequent body of research, Professor Toch has writ‐
ten:
[W]hen used for protection firearms can seriously inhibit aggression and
can provide a psychological buffer against the fear of crime. Furthermore,
the fact that national patterns show little violent crime where guns are
most dense implies that guns do not elicit aggression in any meaningful
way . . . . Quite the contrary, these findings suggest that high saturations
of guns in places, or something correlated with that condition, inhibit
illegal aggression.
Id. at 234 & n.10.
91. KLECK, supra note 8, at 71.
92. See MALCOLM, supra note 10, at 232–33; Alfred Blumstein, Youth Violence, Guns,
and the Illicit‐Drug Industry, 86 J. CRIM. L. & CRIMINOLOGY 10, 21 (1995).
93. See LOTT, supra note 11, at 39 (“[W]hite gun ownership exceed[ed] that for
blacks by about 40 in 1996”); see also KLECK, supra note 8, at 71.
Particularly corrosive to the mantra are the facts as to rural
African‐Americans gun ownership. Per capita, rural African‐
Americans are much more likely to own firearms than are ur‐
ban African‐Americans.94 Yet, despite their greater access to
guns, the firearm murder rate of young rural black males is a
small fraction of the firearm murder rate of young urban black
males.95
These facts are only anomalous in relation to the mantra that
more guns equal more death and fewer guns equal less death.
In contrast, these facts accord with the earlier point regarding
the aberrance of murderers. Whatever their race, ordinary peo‐
ple simply do not murder. Thus preventing law‐abiding, re‐
sponsible African‐Americans from owning guns does nothing
at all to reduce murderers, because they are not the ones who
are doing the killing. The murderers are a small minority of
extreme antisocial aberrants who manage to obtain guns whatever
the level of gun ownership in the African American community.
Indeed, murderers generally fall into a group some crimi‐
nologists have called “violent predators,” sharply differentiat‐
ing them not only from the overall population but from other
criminals as well.96 Surveys of imprisoned felons indicate that
when not imprisoned the ordinary felon averages perhaps 12
crimes per year.97 In contrast, “violent predators” spend much
or most of their time committing crimes, averaging at least 5
assaults, 63 robberies, and 172 burglaries annually.98 A Na‐
tional Institute of Justice survey of 2,000 felons in 10 state pris‐
ons, which focused on gun crime, said of these types of
respondents:
[T]he men we have labeled Predators were clearly omnibus
felons . . . [committing] more or less any crime they had the
opportunity to commit . . . . The Predators (handgun and
shotgun combined) . . . amounted to about 22% of the sam‐
ple and yet accounted for 51% of the total crime [admitted
by the 2,000 felons] . . . . Thus, when we talk about “control‐

94. See LOTT, supra note 11, at 39; see also KLECK, supra note 8, at 71.
95. The murder rate of young urban African Americans is roughly 600% higher
than that of their rural counterparts. See Lois A. Fingerhut et al., Firearm and Nonfire‐
arm Homicide Among Persons 15 Through 19 Years of Age, 267 J. AM. MED. ASS’N 3048,
3049 tbl.1.
96. JAN M. CHAIKEN & MARCIA R. CHAIKEN, VARIETIES OF CRIMINAL BEHAVIOR
62–63 (1982).
97. Id. at 65.
98. Id. at 123, 125, 219 tbl.A.19.
ling crime” in the United States today, we are talking largely
about controlling the behavior of these men.99
The point is not just that demographic patterns of homicide
and gun ownership in the African‐American community do
not support the more guns equal more death mantra. More im‐
portantly, those patterns refute the logic of fewer guns equal
less death. The reason fewer guns among ordinary African‐
Americans does not lead to fewer murders is because that pau‐
city does not translate to fewer guns for the aberrant minority
who do murder. The correlation of very high murder rates with
low gun ownership in African‐American communities simply
does not bear out the notion that disarming the populace as a
whole will disarm and prevent murder by potential murderers.
B. Macro‐historical Evidence: From the
Middle Ages to the 20th Century
The Middle Ages were a time of notoriously brutal and en‐
demic warfare. They also experienced rates of ordinary mur‐
der almost double the highest recorded U.S. murder rate.100
But Middle Age homicide “cannot be explained in terms of
the availability of firearms, which had not yet been in‐
vented.”101 The invention provides some test of the mantra. If
it is true that more guns equal more murder and fewer guns
equal less death, murder should have risen with the inven‐
tion, increased efficiency, and greater availability of firearms
across the population.
Yet, using England as an example, murder rates seem to
have fallen sharply as guns became progressively more effi‐
cient and widely owned during the five centuries after the in‐
vention of firearms.102 During much of this period, because the
entire adult male population of England was deemed to con‐
stitute a militia, every military age male was required to pos‐
sess arms for use in militia training and service.103

99. WRIGHT & ROSSI, supra note 83, at 76.
100. LANE, supra note 65, at 14.
101. Id. at 151. See generally id. ch. 1.
102. MALCOLM, supra note 10, at 19–20.
103. See generally JOYCE LEE MALCOLM, TO KEEP AND BEAR ARMS: THE ORIGINS OF
AN ANGLO‐AMERICAN RIGHT 1–15 (1994); STEPHEN P. HALBROOK, THAT EVERY MAN
BE ARMED: THE EVOLUTION OF A CONSTITUTIONAL RIGHT 37–53 (1984); Don B. Kates,
Handgun Prohibition and the Original Meaning of the Second Amendment, 82 MICH. L.
REV. 204, 214–16 (1983).
The same requirement was true in America during the pe‐
riod of colonial and post‐colonial settlement. Indeed, the basic
English militia laws were superceded by the colonies’ even
more specific and demanding legal requirements of universal
gun ownership. Under those laws, virtually all colonists and
every household were required to own guns. Depending on
the colony’s laws, male youths were deemed of military age at
16, 17, or 18, and every military age man, except for the in‐
sane, infirm, and criminals, had to possess arms. They were
subject to being called for inspection, militia drill, or service,
all of which legally required them to bring and present their
guns. To arm those too poor to afford guns, the laws required
that guns be purchased for them and that they make install‐
ment payments to pay back the cost.104
It bears emphasis that these gun ownership requirements
were not limited to those subject to militia service. Women,
seamen, clergy, and some public officials were automatically
exempt from militia call up, as were men over the upper mili‐

104. MALCOLM, supra note 103, at 138–41; Kates, supra note 102, at 214–16. Typical
laws (quoted with original spelling and punctuation) appear from the following
sources: ARCHIVES OF MARYLAND 77 (William Hand Browne ed., Baltimore, Mary‐
land Historical Society 1883) (“[T]hat every house keeper or housekeepers within
this Province shall have ready continually upon all occasions within his her or their
house for him or themselves and for every person within his her or their house able
to bear armes one Serviceable fixed gunne of bastard muskett boare” along with a
pound of gunpowder, four pounds of pistol or musket shot, “match for match locks
and of flints for firelocks”); NARRATIVES OF EARLY VIRGINIA 273 (Lyon Gardiner
Tyler ed., photo. reprint 1974) (1907) (requiring that everyone attend church on
Sunday, further providing that “all suche as beare armes shall bring their pieces
swordes, poulder and shotte” with them to church on penalty of a fine); RECORDS OF
THE GOVERNOR AND COMPANY OF THE MASSACHUSETTS BAY IN NEW ENGLAND 84
(Nathaniel B. Shurtleff ed., Boston, William White 1853) (ordering towns to provide
their residents with arms if they could not provide their own “for the present, &
after to receive satisfaction for that they disburse when they shall be able”);
RECORDS OF THE COLONY OF RHODE ISLAND AND PROVIDENCE PLANTATIONS, IN
NEW ENGLAND 79‐80, 94 (John Russell Bartlett ed., Providence, A. Crawford Greene
& Brother, 1856) (requiring, respectively: “[T]hat every man do come armed unto
the meeting upon every sixth day,” and also that militia officers go “to every inhabi‐
tant [in Portsmouth and] see whether every one of them has powder” and bullets;
and “that noe man shall go two miles from the Towne unarmed, eyther with Gunn
or Sword; and that none shall come to any public Meeting without his weapon.”);
THE CODE OF 1650, BEING A COMPILATION OF THE EARLIEST LAWS AND ORDERS OF
THE GENERAL COURT OF CONNECTICUT 72 (Hartford, Silas Andrus 1822) (“That all
persons that are above the age of sixteene yeares, except magistrates and church
officers, shall beare arms . . . and every male person within this jurisdiction, above
the said age, shall have in continuall readines, a good muskitt or other gunn, fitt for
service, and allowed by the clark of the band.”).
tary age, which varied from 45 to 60, depending on the col‐
ony. But every household was required to have a gun, even if
all its occupants were otherwise exempt from militia service,
to deter criminals and other attackers. Likewise, all respect‐
able men were theoretically required to carry arms when out
and abroad.105
These laws may not have been fully enforced (except in
times of danger) in areas that had been long‐settled and
peaceful. Nevertheless, “by the eighteenth century, colonial
Americans were the most heavily armed people in the
world.”106 Yet, far from more guns equaling more death, mur‐
ders in the New England colonies were “rare,” and “few”
murderers in all the colonies involved guns “despite their
wide availability.”107
America remained very well armed yet homicide remained
quite low for over two hundred years, from the earliest set‐
tlements through the entire colonial period and early years of
the United States. Homicide in more settled areas only began
rising markedly in the two decades before the Civil War.108 By
that time the universal militia was inoperative and the uni‐
versality of American gun ownership had disappeared as
many people in long‐settled peaceful areas did not hunt and
had no other need for a firearm.109

105. For collections of many of the relevant laws, see Clayton E. Cramer, Gun
Control in Colonial New England, (unpublished manuscript, available at
http://www.claytoncramer.com/GunControlColonialNewEngland.PDF) (last visited
Nov. 19, 2006); Clayton E. Cramer, Gun Control in Colonial New England, Part II
(unpublished manuscript, available at http://www.claytoncramer.com/
GunControlColonialNewEngland2.PDF) (last visited Nov. 19, 2006); Clayton E.
Cramer, Gun Control in the Middle & Southern Colonies, (unpublished manuscript,
available at http://www.claytoncramer.com/MiddleSouthernColonialGunControl.PDF)
(last visited Nov. 19, 2006); Clayton E. Cramer, Militia Statutes,
http://www.claytoncramer.com/primary.html#MilitiaLaws (last visited Nov. 19,
2006).
106. JOHN MORGAN DEDERER, WAR IN AMERICA TO 1775, at 116 (1990).
107. LANE, supra note 64, at 48, 59–60.
108. Id. at 344.
109. The enthusiasm modern gun advocates express for the ancient militia far ex‐
ceeds the enthusiasm felt by the Englishmen and Americans who were actually
subject to the obligations involved. Guns were expensive items even for those own‐
ers who were supplied them by the colonies since they were required to pay the
colonies back over time. And the duty of militia drill was a constant source of irrita‐
tion to men who had little time for leisure and urgent need to devote their time to
making a living for themselves and their families. By the turn of the nineteenth cen‐
tury, at the earliest, the universal militia was in desuetude and replaced in the 1840s

The Civil War acquainted vast numbers of men with mod‐

ern rapid‐fire guns, and, in its aftermath, provided a unique
opportunity to acquire them. Before the Civil War, reliable
multi‐shot rifles or shotguns did not exist and revolvers
(though they had been invented in the 1830s) were so expen‐
sive they were effectively out of reach for most of the Ameri‐
can populace.110 The Civil War changed all that. Officers on
both sides had to buy their own revolvers, while sidearms
were issued to noncommissioned officers generally, as well as
those ordinary soldiers who were in the artillery, cavalry, and
dragoons.111 The fact that over two million men served in the
Union Army at various times while the Confederates had over
half that number suggests the number of revolvers in‐
volved.112

by colorfully garbed volunteer formations whose activities were more social than
military.
110. Revolver inventor Samuel Colt’s first business failed in 1840. It revived itself
only with sales to officers and the military during the Mexican‐American War
(1846–1848), and sustained itself through the 1850s with sales to wealthy Americans
and Europeans. See JOSEPH G. BILBY, CIVIL WAR FIREARMS 157 (1996); LEE KENNETT
& JAMES LAVERNE ANDERSON, THE GUN IN AMERICA 90 (1975); LANE, supra note 65,
at 109. Colt’s sales flourished as foreign armies adopted his revolver and wide sales
took place in the commercial market across Europe, KENNETT & ANDERSON, supra,
at 90, especially after Colt’s prize‐winning exhibit at the 1851 Great Industrial Exhi‐
bition in London. See generally JOSEPH G. ROSA, COLONEL COLT LONDON 13–29
(1976).
111. See generally BILBY, supra note 110, at 157–72. The revolvers involved were by
no means all Colts: “[T]he Federal government also purchased large numbers of
Remington, Starr and Whitney revolvers, as well as the guns of other [American]
makers, including the bizarre looking Savage, with its second ‘ring trigger’ which
cocked the arm, and the sidehammer Joslyn.” Id. at 158. Vast numbers of guns were
also purchased in Europe where, in the first 15 months of the war, the Union bought
over 738,000 firearms (including long arms as well as revolvers). ALLAN R. MILLETT
& PETER MASLOWSKI, FOR THE COMMON DEFENSE: A MILITARY HISTORY OF THE
UNITED STATES OF AMERICA 216 (1984). Some Union infantry units were issued re‐
volvers and many enlisted infantrymen in other units bought their own. BILBY, supra
note 110, at 160.
112. These figures are just estimates. While at least somewhat reliable figures exist
for how many men served at any one time in the Union Army, that number is not
co‐extensive with how many served in total. Some Union soldiers served through‐
out the war, re‐enlisting when their original enlistments were up. Others mustered
out and were replaced with new recruits. Still others deserted long before their
terms were up, again requiring replacements. Some scoundrels enlisted just for the
enlistment bonus, and deserted as soon as they could; some of these went through
the enlistment and desertion process multiple times, collecting a new bonus under a
new name time after time. THE WORLD ALMANAC AND BOOK OF FACTS 2006, at 77
(2006) gives figures of 2,128,948 for the Union Army and 84,415 for the Marines; it
estimates that the Confederate Army’s size was between 600,000 to 1,500,000.
At war’s end, the U.S. Army and Navy were left with vast

numbers of surplus revolvers, both those they had purchased
and those captured from Confederate forces. As the Army
plummeted to slightly over 11,000 men,113 hundreds of thou‐
sands of military surplus revolvers were sold at very low
prices. In addition, when their enlistments were up, or when
they were mustered out at war’s end, former officers and sol‐
diers retained hundreds of thousands of both revolvers and
rifles. These commandeered arms included many of the new
repeating rifles the Union had bought (over the fervent objec‐
tions of short‐sighted military procurement officers) at the
command of President Lincoln, who had tested the Spencer
rifle himself. After his death the Army reverted to the single‐
shot rifle, disposing of all its multi‐shots at surplus and
thereby ruining Spencer by glutting the market.114
Thus over the immediate post‐Civil War years “the country
was awash with military pistols” and rifles of the most mod‐
ern design.115 The final three decades of the century saw the
introduction and marketing of the “two dollar pistol,” which
were very cheap handguns manufactured largely out of pot
metal.116 In addition to being sold locally, such “suicide spe‐
cials” were marketed nationwide through Montgomery Ward
catalogs starting in 1872 and by Sears from 1886.117 They were
priced as low as $1.69, and were marketed under names like
the “Little Giant” and the “Trampʹs Terror.”118

113. RUSSELL F. WEIGLEY, HISTORY OF THE UNITED STATES ARMY 262 (1967) (“The
names of 1,000,516 officers and men were on the [Union Army’s] rolls on May 10,
1865; by [the end of 1866, the draft had ended and] . . . only 11,043 volunteers re‐
mained . . . .”).
114. KENNETT & ANDERSON, supra note 110, at 92–93.
115. DAVID T. COURTWRIGHT, VIOLENT LAND: SINGLE MEN AND SOCIAL DISORDER
FROM THE FRONTIER TO THE INNER CITY 42 (1996).
116. KENNETT & ANDERSON, supra note 110, at 99.
117. Id. at 98–99.
118. Id. at 98–100. An 1879 issue of Scientific American contains an advertisement
for COD purchasing of the $2.75 ”Czar” revolver, presumably an attempt to capital‐
ize on the Smith & Wesson “Russian,” a very high quality weapon that Smith &
Wesson manufactured for the Russian government and sold through the 1870s. SCI.
AM., June 14, 1879, at 381. The 1884 Price List‐Firearms Catalog for N. Curry &
Brother, arms dealers of San Francisco, lists prices from $2.00 for the 7 shot “Fash‐
ion” and “Blue Jacket” revolvers to $2.50 and $3.50 for the “Kitemaug” and
“Ranger” revolvers to various Colt and Smith & Wesson revolvers selling at from
$15.00 to $17.00. KENNETT & ANDERSON, supra note 110, at 98–100.
Thus, the period between 1866 and 1900 saw a vast diffu‐
sion of commercial and military surplus revolvers and lever
action rifles throughout the American populace. Yet, far from
rising, homicide seems to have fallen off sharply during these
thirty years.
Whether or not guns were the cause, homicide steadily de‐
clined over a period of five centuries coincident with the in‐
vention of guns and their diffusion throughout the continent.
In America, from the seventeenth century through the early
nineteenth century, murder was rare and rarely involved
guns, though gun ownership was universal by law and “colo‐
nial Americans were the most heavily armed people in the
world.”119 By the 1840s, gun ownership had declined but
homicide began a spectacular rise through the early 1860s.
From the end of the Civil War to the turn of the twentieth cen‐
tury, however, America in general, and urban areas in par‐
ticular, such as New York, experienced a tremendous spurt in
ownership of higher capacity revolvers and rifles than had
ever existed before, but the number of murders sharply de‐
clined.120
In sum, the notion that more guns equal more death is not
borne out by the historical evidence available for the period
between the Middle Ages and the twentieth century. Yet this
conclusion must be viewed with caution. While one may de‐
scribe broad general trends in murder rates and in the avail‐
ability of firearms, it is not possible to do so with exactitude.
Not until the late 1800s in England, and the mid‐1900s in the
United States were there detailed data on homicide. Informa‐
tion about the distribution of firearms is even more sparse.
For instance, Lane’s generalizations about the rarity of gun
murders and low American murder rates in general are sub‐
ject to some dispute. Professor Randolph Roth, for example,
has shown that early American murder rates and the extent
to which guns were used in murder varied greatly between
differing areas and time periods.121

119. DEDERER, supra note 106, at 116.
120. See LANE, supra note 65, at 181, 307; ERIC H. MONKKONEN, MURDER IN NEW
YORK CITY 21, 30–31, 38 (2001).
121. Randolph Roth, Guns, Gun Culture, and Homicide: The Relationship Between
Firearms, the Uses of Firearms, and Interpersonal Violence, 59 WM. & MARY Q. 223, 234–
40 (2002).
C. Later and More Specific Macro‐Historical Evidence
Malcolm presents reliable trend data on both gun owner‐
ship and crime in England for the period between 1871 and
1964. Significantly, these trend data do not at all correlate as
the mantra would predict: violent crime did not increase with
increased gun ownership nor did it decline in periods in
which gun ownership was lower.122
In the United States, the murder rate doubled in the ten‐
year span between the mid‐1960s and the mid‐1970s. Since
this rise coincided with vastly increasing gun sales, it was
viewed by many as proof positive that more guns equal more
death. That conclusion, however, does not follow. It is at least
equally possible that the causation was reversed: that is, the
decade’s spectacular increases in murder, burglary, and all
kinds of violent crimes caused fearful people to buy guns.123
The dubiousness of assuming that the gun sales caused the
rise in murder rather than the reverse might have been clearer
had it been known in this period that virtually the same mur‐
der rate increase was occurring in gun‐less Russia.124 Clearly
there is little basis to assume guns were the reason for the
American murder rate rise when the Russian murder rate ex‐
hibited the same increase without a similar increase in the
number of guns.
Reliable information on both gun ownership and murder
rates in the United States is available only for the period
commencing at the end of World War II. Significantly, the
decade from the mid‐1960s to the mid‐1970s is a unique ex‐
ception to the general pattern that, decade‐by‐decade, the
number of guns owned by civilians has risen steadily and

122. See MALCOLM, supra note 10, app. at 258. The handgun ownership data cited
are tax data and so doubtless fail to count the pistols owned by criminals and others
who failed to pay taxes. The extremely low numbers of gun crimes, however, do not
support the notion that there were numerous criminal owners of guns, or at least
that they used the guns for crime.
123. In contrast to the more guns equal more death mantra, studies suggest that
crime rate increases fuel gun buying, rather than the other way around. See, e.g.,
Douglas C. Bice & David D. Hemley, The Market for New Handguns: An Empirical
Investigation, 45 J.L. & ECON. 251, 253, 261–262 (2002); Lawrence Southwick, Jr., Do
Guns Cause Crime? Does Crime Cause Guns? A Granger Test, 25 ATLANTIC ECON. J. 256,
256, 272 (1997); KLECK, supra note 8, at 79–81.
124. In 1965, the Russian homicide rate stood at 5.9 per 100,000 population while
the American rate was 5.4. As of 1975, both Russian and American rates had nearly
doubled, the Russian to 10.3 and the American to 9.7. See Pridemore, supra note 2, at
272 fig.2; see also supra note 6 and accompanying text.
dramatically but murder rates nevertheless have remained

stable or even declined. As for the second half of the twenti‐
eth century, and especially its last quarter, a study compar‐
ing the number of guns to murder rates found that during
the 25‐year period from 1973 to 1997, the number of hand‐
guns owned by Americans increased 160% while the number
of all firearms rose 103%. Yet over that period, the murder
rate declined 27.7%.125 It continued to decline in the years
1998, 1999, and 2000, despite the addition in each year of two
to three million handguns and approximately five million
firearms of all kinds. By the end of 2000, the total American
gunstock stood at well over 260 million—951.1 guns for
every 1,000 Americans—but the murder rate had returned to
the comparatively low level prior to the increases of the mid‐
1960s to mid‐1970s period.126
In sum, the data for the decades since the end of World War
II also fails to bear out the more guns equal more death man‐
tra. The per capita accumulated stock of guns has increased,
yet there has been no correspondingly consistent increase in
either total violence or gun violence. The evidence is consis‐
tent with the hypothesis that gun possession levels have little
impact on violence rates.127
D. Geographic Patterns within Nations
Once again, if more guns equal more death and fewer guns
equal less death, areas within nations with higher gun owner‐
ship should in general have more murders than those with
less gun ownership in a similar area. But, in fact, the reverse
pattern prevails in Canada,128 “England, America, and Swit‐
zerland, [where the areas] with the highest rates of gun own‐
ership were in fact those with the lowest rates of violence.”129

125. Don B. Kates & Daniel D. Polsby, Long‐Term Nonrelationship of Widespread and
Increasing Firearm Availability to Homicide in the United States, 4 HOMICIDE STUD. 185,
190–91 (2000).
126. See communication from Gary Kleck, Professor, Florida State University, to
Don B. Kates and Gary Mauser (Feb. 26, 2003) (on file with Authors).
127. KLECK, supra note 8, at 17–19.
128. See Philip C. Stenning, Gun Control: A Critique of Current Policy, POL’Y
OPTIONS, Oct. 1994, at 13, 15.
129. MALCOLM, supra note 10, at 204; see also BBC News, Handgun Crime ‘Up’ De‐
spite Ban, July 16, 2001, http://news.bbc.co.uk/2/hi/uk_news/1440764.stm (noting that
English areas with very low numbers of firearms have higher than average gun
crime while areas with the highest levels of legally held guns do not).
A recent study of all counties in the United States has again
demonstrated the lack of relationship between the prevalence
of firearms and homicide.130
This inverse correlation is one of several that seems to
contradict more guns equal more death. For decades the
gun lobby has emphasized that, in general, the American
jurisdictions where guns are most restricted have consis‐
tently had the highest violent crime rates, and those with
the fewest restrictions have the lowest violent crime rates.131
For instance, robbery is highest in jurisdictions which are
most restrictive of gun ownership.132As to one specific con‐
trol, the ban on carrying concealed weapons for protection,
“violent‐crime rates were highest in states [that flatly ban
carrying concealed weapons], next highest in those that al‐
lowed local authorities discretion [to deny] permits, and
lowest in states with nondiscretionary” concealed weapons
laws under which police are legally required to license
every qualified applicant.133 Also of interest are the exten‐
sive opinion surveys of incarcerated felons, both juvenile
and adult, in which large percentages of the felons replied
that they often feared potential victims might be armed and
aborted violent crimes because of that fear.134 The felons
most frightened about confronting an armed victim were
those “from states with the greatest relative number of pri‐
vately owned firearms.”135

130. Tomislav Kovandzic, Mark E. Schaffer, & Gary Kleck, Gun Prevalence, Homi‐
cide Rates and Causality: A GMM Approach to Endogeneity Bias 39–40 (Ctr. for Econ.
Policy Research, Discussion Paper No. 5357, 2005) available at
http://ssrn.com/abstract=878132.
131. See e.g., National Rifle Association Institute for Legislative Action (NRA‐ILA),
More Guns, Less Crime (Sept. 26, 2006), http://www.nraila.org/Issues/
FactSheets/Read.aspx?idequals206&issueequals007; NRA‐ILA, The War Against
Handguns (Feb. 15, 2001), www.nraila.org/Issues/FactSheets/Read.aspx?idequals17;
NRA‐ILA, Right‐to‐Carry 2006 (Oct. 3, 2006), www.nraila.org/Issues/FactSheets/
Read.aspx?idequals18; NRA‐ILA, The Brady Handgun Violence Prevention Act,
Does it Live Up to its Name? (July 28, 1999), www.nraila.org/Issues/FactSheets/
Read.aspx?idequals73.
132. See, e.g., Philip J. Cook, The Effect of Gun Availability on Robbery and Robbery
Murder: A Cross Section Study of Fifty Cities, 3 POL’Y STUD. REV. ANN. 743, 770 (1979).
133. LOTT, supra note 11, at 43.
134. WRIGHT & ROSSI, supra note 83, at 147, 150.
135. Id. at 151.
E. Geographic Comparisons: European Gun
Ownership and Murder Rates
This topic has already been addressed at some length in connec‐
tion with Tables 1–3, which contain the latest data available. Tables
4–6, contain further, and somewhat more comprehensive, data
from the early and mid‐1990s.136 These statistics reinforce the point
that murder rates are determined by basic socio‐cultural and eco‐
nomic factors rather than mere availability of some particular form
of weaponry. Consider Norway and its neighbors Sweden, the
Netherlands, and Denmark. Norway has far and away Western
Europe’s highest household gun ownership rate (32%), but also its
lowest murder rate. The Netherlands has the lowest gun owner‐
ship rate in Western Europe (1.9%), and Sweden lies midway be‐
tween (15.1%) the Netherlands and Norway. Yet the Dutch gun
murder rate is higher than the Norwegian, and the Swedish rate is
even higher, though only slightly.137
Table 4: Intentional Deaths: United States vs.
Continental Europe Rates
In order of highest combined rate; nations having higher rates than the
United States are indicated by asterisk (suicide rate) or + sign (murder rate).
Nation Suicide Murder Combined rates

Russia 41.2* 30.6+ 71.8
Estonia 40.1* 22.2+ 62.3
Latvia 40.7* 18.2+ 58.9
Lithuania 45.6* 11.7+ 57.3
Belarus 27.9* 10.4+ 38.3
Hungary 32.9* 3.5 36.4
Ukraine 22.5* 11.3+ 33.8
Slovenia 28.4* 2.4 30.4
Finland 27.2* 2.9 30.1
Denmark 22.3* 4.9 27.2
Croatia 22.8* 3.3 26.1
Austria 22.2* 1.0 23.2
Bulgaria 17.3* 5.1 22.4
France 20.8* 1.1 21.9
Switzerland 21.4* 1.1‡ 24.1

136. Tables 4–6 were previously published as appendices to Kates, supra note 81,
app. at 81 tbl.1, 82 tbl.2, 83 tbl.3.
137. See infra Table 5.
Belgium 18.7* 1.7 20.4

United States 11.6 7.8 19.4
Poland 14.2* 2.8 17.0
Germany 15.8* 1.1 16.9
Romania 12.3* 4.1 16.4
Sweden 15.3* 1.0 16.3
Norway 12.3* 0.8 13.1
Holland 9.8 1.2 11.0
Italy 8.2 1.7 9.9
Portugal 8.2 1.7 9.9
Spain 8.1 0.9 9.0
Greece 3.3 1.3 4.6
Notes: Data based in general on U.N. DEMOGRAPHIC YEARBOOK (1998) as re‐
ported in David C. Stolinsky, America: The Most Violent Nation? 5 MED. SENTINEL
199–201 (2000). It should be understood that, though the 1998 Yearbook gives fig‐
ures for as late as 1996, the figures are not necessarily for that year. The Yearbook
contains the latest figure each nation has provided the U.N., which may be 1996,
1995, or 1994.
‡
The Swiss homicide figure that Stolinsky reports is an error because it com‐
bines attempts with actual murders. We have computed the Swiss murder rate
by averaging the 1994 and 1995 Swiss National Police figures for actual murders
in those years given in RICHARD MUNDAY & JAN A. STEVENSON, GUNS AND
VIOLENCE: THE DEBATE BEFORE LORD CULLEN 268 (1996).
Table 5: European Gun/Handgun Violent Death
Percent of
Suicide Murder Percent of households
with with households with hand‐
Nation handgun handgun with guns guns
Belgium 18.7 1.7 16.6% 6.8%
France 20.8 1.1 22.6% 5.5%
West Germany 15.8 1.1 8.9% 6.7%*
Holland 9.8 1.2 1.9% 1.2%
Italy 8.2 1.7 16.0% 5.5%
Norway 12.3 0.8 32%% 3.8%
Sweden 15.3 1.3 15.1% 1.5%
Switzerland 20.8 1.1** 27.2% 12.2%
Notes: For derivation of the homicide rates, see notes to Table 4. The data
on household firearms ownership come from British Home Office figures
printed in RICHARD MUNDAY & JAN A. STEVENSON, GUNS AND VIOLENCE: THE
DEBATE BEFORE LORD CULLEN 30, 275 (1996).
* Note that the data here are for West Germany and were obtained when
that nation still existed as an independent entity. See infra Tables 1 & 4 for later
(but differently derived) data for the current nation of Germany.
** Again, the Swiss homicide figure that Stolinsky reports is an error be‐
cause it combines attempts with actual murders. See notes for Table 4.
Table 6: European Firearms‐Violent Deaths
Number of
Suicide Murder guns per
with with 100,000
Nation Suicide gun Murder gun population
Austria N/A N/A 2.14 0.53 41.02*
Belarus 27.26 N/A 9.86 N/A 16.5
Czech Rep. 9.88 1.01 2.80 0.92 27.58
Estonia 39.99 3.63 22.11 6.2 28.56
Finland 27.28 5.78 3.25 0.87 411.20**
Germany 15.80 1.23 1.81 0.21 122.56
Greece 3.54 1.30 1.33 0.55 77.00
Hungary 33.34 0.88 4.07 0.47 15.54
Moldova N/A N/A 17.06 0.63 6.61
Poland 14.23 0.16 2.61 0.27 5.30
Romania N/A N/A 4.32 0.12 2.97
Slovakia 13.24 0.58 2.38 0.36 31.91
Spain 5.92 N/A 1.58 0.19 64.69
Sweden 15.65 1.95 1.35 0.31 246.65
Notes: It bears emphasis that the following data come from a special U.N.
report whose data are not fully comparable to those in Tables 4 and 5 because
they cover different years and derive from substantially differing sources.138
This special report is based on data obtained from the governments of the na‐
tions set out below, especially data on gun permits or other official indicia of
gun ownership in those nations.139 The data on suicide and murder in those

138. The data derive from a much more extensive survey of legal firearms owner‐
ship in numerous nations which was carried out by researchers provided by the
Government of Canada under the auspices of the United Nations Economic and
Social Council, Commission on Crime Prevention and Criminal Justice in 1997. The
entire survey is published as a report to the Secretary General on April 25, 1997 as
E/CN.15/1997/4. That report is analysed in some detail in an unpublished paper (“A
Cross Sectional Study of the Relationship Between Levels of Gun Ownership and
Violent Deaths”) written by the leading English student of firearms regulation, re‐
tired Chief Superintendent of English police Colin Greenwood of the Firearms Re‐
search and Advisory Service. We are indebted to Chief Superintendent Greenwood
for the opportunity to review his paper. Note that in the table which follows we
have focused only on European nations.
139. The gun ownership data in Table 4 derive from a random telephone survey
on gun ownership in various nations. Chief Superintendent Greenwood’s paper is
contemptuous of such data, in part because people may be unwilling to acknowl‐
edge owning guns to telephoning pollsters. For similar doubts see Don B. Kates &
Daniel D. Polsby, Long Term Non‐Relationship of Firearm Availability to Homicide, 4
HOMICIDE STUD. 185–201 (2000). But that was in the context of comparing survey
data on the number of guns owned to production and important data that are un‐
questionably more comprehensive and superior in every way. Chief Superintendent
Greenwood himself admits that the special U.N. report data are not necessarily

nations also come from their governments as do the similar data in Tables 4
and 5, but for later years, and also include data on the number of firearm
homicides and firearm suicides which are not available from the U.N. source
used in Tables 4 and 5.
* This may well be an undercount because an Austrian license is not limited
to a single firearm but rather allows the licensee to possess multiple guns.
** The source from which Table 5 derives also gives figures for Finland,
which we have omitted there because they are earlier and closely similar ex‐
cept in one respect: instead of official ownership figures for guns, they give a
survey‐based figure for households having a gun: 23.2%.

These comparisons are reinforced by Table 6, which gives differ‐
ently derived (and non‐comparable) gun ownership rates, overall
murder rates, and rates of gun murder, for a larger set of European
nations.140 Table 6 reveals that even though Sweden has more than
double the rate of gun ownership as neighboring Germany, as well
as more gun murders, it has 25% less murder overall. In turn, Ger‐
many, with three times the gun ownership rate of neighboring Aus‐
tria, has a substantially lower murder rate overall and a lower gun
murder rate. Likewise, though Greece has over twice the per capita
gun ownership rate of the Czech Republic, Greece has substantially
less gun murder and less than half as much murder overall. Al‐
though Spain has over 12 times more gun ownership than Poland,
the latter has almost a third more gun murder and more overall
murder than the former. Finally, Finland has 14 times more gun
ownership than neighboring Estonia, yet Estonia’s gun murder and
overall murder rates are about seven times higher than Finland’s.
F. Geographic Comparisons: Gun Ownership and Suicide Rates
The mantra more guns equal more death and fewer guns equal
less death is also used to argue that “limiting access to firearms
could prevent many suicides.”141 Once again, this assertion is di‐

comprehensive and are problematic in various other respects. Even assuming they
are clearly superior to the survey data, the latter cover multiple nations that the
special U.N. report does not. Given that neither source is indubitable, it seems pref‐
erable to have such information on those nations as the survey data reveal, rather
than no data at all.
140. Table 6 covers different years from Table 5, its comparative gun ownership
figures derive from government records rather than survey data, and it gives rates
for gun murders, data that are not available in the sources from which Table 5 is
taken. See the explanatory note that precedes Table 6.
141. Arthur L. Kellermann et al., Suicide in the Home in Relation to Gun Ownership,
327 NEW ENG. J. MED. 467, 467, 471–72 (1992); see also Antoon Leenaars, et al., Con‐
trolling the Environment to Prevent Suicide: International Perspectives, 45 CAN. J.
PSYCHIATRY 639 (2000).
rectly contradicted by the studies of 36 and 21 nations (respec‐
tively) which find no statistical relationship. Overall suicide rates
were no worse in nations with many firearms than in those where
firearms were far less widespread.142
Consider the data about European nations in Tables 5 and 6.
Sweden, with over twice as much gun ownership as neighboring
Germany and a third more gun suicide, nevertheless has the lower
overall suicide rate. Greece has nearly three times more gun own‐
ership than the Czech Republic and somewhat more gun suicide,
yet the overall Czech suicide rate is over 175% higher than the
Greek rate. Spain has over 12 times more gun ownership than Po‐
land, yet the latter’s overall suicide rate is more than double the
former’s. Tragically, Finland has over 14 times more gun owner‐
ship than neighboring Estonia, and a great deal more gun‐related
suicide. Estonia, however, turns out to have a much higher suicide
rate than Finland overall.
There is simply no relationship evident between the extent of
suicide and the extent of gun ownership. People do not commit
suicide because they have guns available. In the absence of fire‐
arms, people who are inclined to commit suicide kill themselves
some other way.143 Two examples seem as pertinent as they are
poignant. The first concerns the 1980s increase in suicide among
young American males, an increase that, although relatively mod‐
est, inspired perfervid denunciations of gun ownership.144 What
these denunciations failed to mention was that suicide of teenagers
and young adults was increasing throughout the entire industrial‐
ized world, regardless of gun availability, and often much more
rapidly than in the United States. The only unusual aspect of sui‐
cide in the United States was that it involved guns. The irrelevancy
of guns to the increase in American suicide is evident because sui‐
cide among English youth actually increased 10 times more

142. See Killias et al., supra note 42, at 430 (study of 21 nations); see generally KLECK,
supra note 8.
143. See KLECK, supra note 8, at ch. 8; see also World Health Organization, supra
note 43, at 3 (showing that around the world “firearms accounted for only one‐fifth
of all suicides, just ahead of poisoning . . . . [s]trangulation, i.e. (hanging) was the
most frequently used method of suicide”).
144. See, e.g., Jeffrey H. Boyd & Eve K. Moscicki, Firearms and Youth Suicide, 76 AM.
J. PUB. HEALTH 1240 (1986); James A. Mercy et al., Public Health Policy for Preventing
Violence, 12 HEALTH AFF. 7 (1993); Daniel W. Webster & Modena E. H. Wilson, Gun
Violence Among Youth and the Pediatricianʹs Role in Primary Prevention, 94 PEDIATRICS
617 (1994); Lois A. Fingerhut & Joel C. Kleinman, Firearm Mortality Among Children
and Youth, NAT’L CTR HEALTH STAT. ADVANCE DATA, Nov. 3, 1989, at 1.
sharply, with “car exhaust poisoning [being] the method of suicide
used most often.”145 By omitting such facts, the articles blaming
guns for increasing American suicide evaded the inconvenience of
having to explain exactly what social benefit nations with few guns
received from having their youth suicides occur in other ways.
Even more poignant are the suicides of many young Indian
women born and raised on the island of Fiji. In general, women
are much less likely to commit suicide than are men.146 This sta‐
tistic is true of Fijian women overall as well, but not of women
in the large part of Fiji’s population that is of Indian ancestry.
As children, these Indian women are raised in more‐or‐less lov‐
ing and supportive homes. But upon marriage they are dis‐
persed across the island to remote areas where they live with
their husbands’ families, an often overtly hostile situation the
husbands do little to mitigate. Indian women on Fiji have a sui‐
cide rate nearly as high as that of Indian men, a rate many
times greater than that of non‐Indian Fijian women.147 It also
bears emphasis that the overall Fijian suicide rate far exceeds
that of the United States.
The method of suicide is particularly significant. Fijian
women of Indian ancestry commit suicide without using
guns, perhaps because guns are unavailable. About three‐
quarters of these women hang themselves, while virtually all
the rest die from consuming the agricultural pesticide
paraquat. The recommendation of the author whose article
chronicles all these suicides is so myopic as to almost carica‐
ture the more guns equal more death mindset: to reduce sui‐
cide by Indian women, she recommends that the Fijian state
stringently control paraquat.148 Apparently she believes de‐

145. Keith Hawton, By Their Own Young Hand, 304 BRIT. MED. J. 1000 (1992); see
also Teenage Deaths Increasing Across Europe, CRIM. & JUST. INT’L, Nov.–Dec. 1991,
at 4.
146. World Health Organization, Suicide Rates by Country,
http://www.who.int/mental_health/prevention/suicide/country_reports/en/
index.html (follow hyperlinks to specific countries) (last visited Jan. 18, 2007). For
example, in the United States, suicide rates for males exceed those for females by a
17.9‐4.2 margin (2002 data). In Denmark, the margin is 19.2‐8.1 (2001 data); in Aus‐
tria, the margin is 27.0‐8.2 (2004 data); and in Belgium, the margin is 31.2‐11.4 (1997
data).
147. See Ruth H. Haynes, Suicide in Fiji: A Preliminary Study, 145 BRIT. J.
PSYCHIATRY 433 (1984).
148. Id. at 437. More or less the same situation seems to prevail in the substantially
Indian‐populated nation of Sri Lanka (formerly Ceylon). It “has one of the highest
suicide rates in the world . . . . Suicides are especially frequent among young adults,

creased access to a means of death will reconcile these

women to a life situation they regard as unendurable. At the
risk of belaboring what should be all too obvious, restricting
paraquat will not improve the lives of these poor women. It
will only reorient them towards hanging, drowning, or some
other means of suicide.
Guns are just one among numerous available deadly instru‐
ments. Thus, banning guns cannot reduce the amount of sui‐
cides. Such measures only reduce the number of suicides by
firearms. Suicides committed in other ways increase to make
up the difference. People do not commit suicide because they
have guns available. They kill themselves for reasons they
deem sufficient, and in the absence of firearms they just kill
themselves in some other way.
CONCLUSION
This Article has reviewed a significant amount of evidence
from a wide variety of international sources. Each individual
portion of evidence is subject to cavil—at the very least the
general objection that the persuasiveness of social scientific
evidence cannot remotely approach the persuasiveness of
conclusions in the physical sciences. Nevertheless, the bur‐
den of proof rests on the proponents of the more guns equal
more death and fewer guns equal less death mantra, espe‐
cially since they argue public policy ought to be based on
that mantra.149 To bear that burden would at the very least
require showing that a large number of nations with more
guns have more death and that nations that have imposed
stringent gun controls have achieved substantial reductions
in criminal violence (or suicide). But those correlations are
not observed when a large number of nations are compared
across the world.

both male and female. Compared to the U.S., the suicide rate for males ages 15 to 24
years in Sri Lanka is nearly four times greater; the female rate nearly 13 times
greater. The most common mode of suicide is ingestion of liquid pesticides.” Lawrence R.
Berger, Suicides and Pesticides in Sri Lanka, 78 AM. J. PUB. HEALTH 826 (1988) (empha‐
sis added).
149. (1) Those who propose to change the status quo bear the burden of proving
that change is a good idea; (2) those who propose a new policy bear the burden of
proving that the policy is a good idea; and (3) in a free society those who propose to
abolish a personal liberty passionately valued by millions bear the burden of prov‐
ing that abolishment is a good idea.
Over a decade ago, Professor Brandon Centerwall of the Uni‐

versity of Washington undertook an extensive, statistically sophis‐
ticated study comparing areas in the United States and Canada to
determine whether Canada’s more restrictive policies had better
contained criminal violence. When he published his results it was
with the admonition:
If you are surprised by [our] finding[s], so [are we]. [We] did
not begin this research with any intent to “exonerate” hand‐
guns, but there it is—a negative finding, to be sure, but a nega‐
tive finding is nevertheless a positive contribution. It directs us
where not to aim public health resources.150

150. Brandon S. Centerwall, Authorʹs Response to “Invited Commentary: Common
Wisdom and Plain Truth,” 134 AM. J. EPIDEMIOLOGY 1264, 1264 (1991).

Harvard - Would Banning Firearms Reduce Murders and Suicide

Uploaded by

Copyright:

Available Formats

Harvard - Would Banning Firearms Reduce Murders and Suicide

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Harvard - Would Banning Firearms Reduce Murders and Suicide

Uploaded by

Copyright:

Available Formats

Nation Murder Rate Rate of Gun Ownership

nation’s nineteenth and twentieth century experience as fol‐

perience deserves note, given how often and favorably advo‐

To conserve the resources of the inundated criminal justice

1980s and 1990s. During these two decades, while Britain and

However unintentionally, the irrelevance of focusing on

prove gun advocates’ assertion that gun controls actually

Nation Handgun Policy Murder Rate Year

murder rates than neighboring nations that permit handgun

The “more guns equal more death” mantra seems plausible

tion, thereby deterring Norwegian criminals from committing

Nation Murder Rate Rate of Gun Ownership

The Civil War acquainted vast numbers of men with mod‐

At war’s end, the U.S. Army and Navy were left with vast

dramatically but murder rates nevertheless have remained

Nation Suicide Murder Combined rates

Belgium 18.7* 1.7 20.4

creased access to a means of death will reconcile these

Over a decade ago, Professor Brandon Centerwall of the Uni‐

You might also like