Gun Buy-Backs
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
102 103 104 105 106 107 108 109 110 111 112 113 114 115
Gun buy-back programs involve a government or private group paying
individuals to turn in guns they possess. The programs do not require the
participants to identify themselves, in order to encourage participation by
offenders or those with weapons used in crimes. The guns are then destroyed.
The theoretical premise for gun buy-back programs is that the
program will lead to fewer guns on the streets because fewer guns are
available for either theft or trade, and that consequently violence will decline.
It is the committee’s view that the theory underlying gun buy-back
programs is badly flawed and the empirical evidence demonstrates the
ineffectiveness of these programs.
The theory on which gun buy-back programs is based is flawed in three
respects. First, the guns that are typically surrendered in gun buy-backs are
those that are least likely to be used in criminal activities. Typically, the
guns turned in tend to be of two types: (1) old, malfunctioning guns whose
resale value is less than the reward offered in buy-back programs or (2)
guns owned by individuals who derive little value from the possession of
the guns (e.g., those who have inherited guns). The Police Executive Research
Forum (1996) found this in their analysis of the differences between
weapons handed in and those used in crimes. In contrast, those who are
either using guns to carry out crimes or as protection in the course of
engaging in other illegal activities, such as drug selling, have actively acquired
their guns and are unlikely to want to participate in such programs.
Second, because replacement guns are relatively easily obtained, the
actual decline in the number of guns on the street may be smaller than the
number of guns that are turned in. Third, the likelihood that any particular
gun will be used in a crime in a given year is low. In 1999, approximately
6,500 homicides were committed with handguns. There are approximately
70 million handguns in the United States. Thus, if a different handgun were
used in each homicide, the likelihood that a particular handgun would be
used to kill an individual in a particular year is 1 in 10,000. The typical gun
buy-back program yields less than 1,000 guns. Even ignoring the first two
points made above (the guns turned in are unlikely to be used by criminals
and may be replaced by purchases of new guns), one would expect a reduction
of less than one-tenth of one homicide per year in response to such a
gun buy-back program. The program might be cost-effective if those were
the correct parameters, but the small scale makes it highly unlikely that its
effects would be detected.
In light of the weakness in the theory underlying gun buy-backs, it is
not surprising that research evaluations of U.S. efforts have consistently
failed to document any link between such programs and reductions in gun
violence (Callahan et al., 1994; Police Executive Research Forum, 1996;
Rosenfeld, 1996).
Outside the United States there have been a small number of buy-backs
of much larger quantities of weapons, in response to high-profile mass
murders with firearms. Following a killing of 35 persons in Tasmania in
1996 by a lone gunman, the Australian government prohibited certain
categories of long guns and provided funds to buy back all such weapons in
private hands (Reuter and Mouzos, 2003). A total of 640,000 weapons
were handed in to the government (at an average price of approximately
$350), constituting about 20 percent of the estimated stock of weapons.
The weapons subject to the buy-back, however, accounted for a modest
share of all homicides or violent crimes more generally prior to the buyback.
Unsurprisingly, Reuter and Mouzos (2003) were unable to find evidence
of a substantial decline in rates for these crimes. They noted that in
the six years following the buy-back, there were no mass murders with
firearms and fewer mass murders than in the previous period; these are
both weak tests given the small numbers of such incidents annually.
Banning Assault Weapons
In 1994, Congress enacted the Violent Crime Control and Law Enforcement
Act, which banned the importation and manufacture of certain
military-style semiautomatic “assault” weapons and ammunition magazines
capable of holding more than 10 rounds (National Institute of Justice,
1997). Assault weapons and large-capacity magazines manufactured before
the effective date of the ban were grandfathered and thus legal to own and
transfer. These guns are believed to be particularly dangerous because they
facilitate the rapid firing of high numbers of shots. While assault weapons
and large-capacity magazines are used only in a modest fraction of gun
crimes, the premise of the ban was that a decrease in their use may reduce
gunshot victimization, particularly victimizations involving multiple wounds
or multiple victims (Roth and Koper, 1997).
A recent evaluation of the short-term effects of the 1994 federal assault
weapons ban did not reveal any clear impacts on gun violence outcomes
(Koper and Roth, 2001b). Using state-level Uniform Crime Reports data on
gun homicides, the authors of this study suggest that the potential impact of
the law on gun violence was limited by the continuing availability of assault
weapons through the ban’s grandfathering provision and the relative rarity
with which the banned guns were used in crime before the ban. Indeed, as
the authors concede and other critics suggest (e.g., Kleck, 2001), given the
nature of the intervention, the maximum potential effect of the ban on gun
violence outcomes would be very small and, if there were any observable
effects, very difficult to disentangle from chance yearly variation and other
state and local gun violence initiatives that took place simultaneously. In a
subsequent paper on the effects of the assault weapons ban on gun markets,
Koper and Roth (2001a) found that, in the short term, the prices of assault
weapons in both primary and legal secondary markets rose substantially at
the time of the ban, and this may have reduced the availability of the assault
weapons to criminals. However, this increase in price was short-lived as a
surge in assault weapon production in the months prior to the ban and the
availability of legal substitutes caused prices to fall back to nearly preban
levels. The ban is also weakened by the ease with which legally available
guns and magazines can be altered to evade the intent of the ban. The
results of these two studies should be interpreted with caution, since any
trends observed in the relatively short study time period (24-month followup
period) are unlikely to predict long-term trends accurately.
District of Columbia Handgun Ban
Bans on the ownership, possession, or purchase of guns are the most
direct means available to policy makers for reducing the prevalence of guns.
The District of Columbia’s Firearms Control Regulations Act of 1975 is the
most carefully analyzed example of a handgun ban. This law prohibited the
purchase, sale, transfer, and possession of handguns by D.C. residents other
than law enforcement officers or members of the military. Note, however,
that individuals who had previously registered handguns prior to the passage
of this law were allowed to keep them under this law. Long guns were
not covered by the ban.6
One would expect the passage of the District’s handgun ban to have
little impact on the existing stock of legally held handguns but to greatly
reduce the flow of new handguns to law-abiding citizens. Over time, the
number of legally held handguns will decline. It is less clear how the illegal
6For a more detailed discussion of the law and the politics surrounding its passage, see
Jones (1981).
possession of guns will be affected. The flow of new guns to the illegal
sector may be reduced to the extent that legal guns enter the illegal sector
through resale or theft from the legal stock in the District. Theory alone
cannot determine whether this handgun ban will reduce crime and violence
overall. One would expect that the share of crimes in which guns are used
should decline over time if the handgun ban is effective.
The empirical evidence as to the success of the Washington, DC, handgun
ban is mixed. Loftin et al. (1991) used an interrupted-time-series methodology
to analyze homicides and suicides in Washington, DC, and the
surrounding areas of Maryland and Virginia before and after the introduction
of the ban. They included the suburban areas around Washington, DC,
as a control group, since the law does not directly affect these areas. Using
a sample window of 1968-1987, they report a 25 percent reduction in gunrelated
homicides in the District of Columbia after the handgun ban and a
23 percent reduction in gun-related suicides. In contrast, the surrounding
areas of Maryland and Virginia show no consistent patterns, suggesting a
possible causal link between the handgun ban and the declines in gunrelated
homicide and suicide. In addition, Loftin et al. (1991) report that
nongun-related homicides and suicides declined only slightly after the handgun
ban, arguing that this is evidence against substitution away from guns
toward other weapons.
Britt et al. (1996), however, demonstrate that the earlier conclusions of
Loftin et al. (1991) are sensitive to a number of modeling choices. They
demonstrate that the same handgun-related homicide declines observed in
Washington, DC, also occurred in Baltimore, even though Baltimore did
not experience any change in handgun laws.7 Thus, if Baltimore is used as
a control group rather than the suburban areas surrounding DC, the conclusion
that the handgun law lowered homicide and suicide rates does not
hold. Britt et al. (1996) also found that extending the sample frame an
additional two years (1968-1989) eliminated any measured impact of the
handgun ban in the District of Columbia. Furthermore, Jones (1981) discusses
a number of contemporaneous policy interventions that took place
around the time of the Washington, DC, gun ban, which further call into
question a causal interpretation of the results.
In summary, the District of Columbia handgun ban yields no conclusive
evidence with respect to the impact of such bans on crime and violence. The
nature of the intervention—limited to a single city, nonexperimental, and
accompanied by other changes that could also affect handgun homicide—
make it a weak experimental design. Given the sensitivity of the results to
alternative specifications, it is difficult to draw any causal inferences.
7Britt et al. (1996) do not report results for suicide.