Impacts of the 1994 Assault Weapons

Home , Assault weapon, Automatic firearm, Closed bolt, Pistol grip, Steyr AUG

NT OF ME J T US U.S. Department of Justice R T A I P C E E D

O J C S Office of Justice Programs F A V M F O I N A C I J S R E BJ G O OJJ DP O F PR National Institute of Justice JUSTICE National Institute of Justice R e s e a r c h i n B r i e f Jeremy Travis, Director March 1999 Issues and Findings Impacts of the 1994 Assault Discussed in this Brief: This study examines the short-term impact Weapons Ban: 1994–96 (1994–96) of the assault weapons ban on gun markets and gun- by Jeffrey A. Roth and Christopher S. Koper related violence as contained in Title XI of the Federal Violent Crime On January 17, 1989, Patrick Edward zines. The legislation required the Attor- Control and Law Enforcement Act Purdy, armed with an AKS rifle—a ney General to deliver to Congress within of 1994. Title XI prohibits the semiautomatic variant of the military 30 months an evaluation of the effects of manufacture, sale, and possession AK–47—returned to his childhood the ban. To meet this requirement, the of specific makes and models of military-style semiautomatic fire- elementary school in Stockton, California, National Institute of Justice (NIJ) funded arms and other semiautomatics and opened fire, killing 5 children and research from October 1995 to December with multiple military-style features wounding 30 others. Purdy, a drifter, 1996 to evaluate the impact of Subtitle A. (detachable magazines, flash sup- squeezed off more than 100 rounds in This Research in Brief summarizes the pressors, folding rifle stocks, and 1 minute before turning the weapon on results of that evaluation. threaded barrels for attaching himself. silencers) and outlaws most large A number of factors—including the fact capacity magazines (ammunition- During the 1980s and early 1990s, this that the banned weapons and magazines feeding devices) capable of holding tragedy and other similar acts of seem- were rarely used to commit murders in more than 10 rounds of ammuni- ingly senseless violence, coupled with this country, the limited availability of tion. Weapons and magazines escalating turf and drug wars waged by data on the weapons, other components of manufactured prior to September urban gangs, sparked a national debate the Crime Control Act of 1994, and State 13, 1994, are exempt from the ban. over whether legislation was needed to and local initiatives implemented at the Key issues: Although the weapons end, or at least restrict, the market for im- same time—posed challenges in discern- banned by this legislation were used ported and domestic “assault weapons.” ing the effects of the ban. The ban ap- only rarely in gun crimes before Beginning in 1989, a few States enacted pears to have had clear short-term effects the ban, supporters felt that these their own assault weapons bans, but it on the gun market, some of which were weapons posed a threat to public was not until 1994 that a Federal law was unintended consequences: production of safety because they are capable of enacted. the banned weapons increased before the firing many shots rapidly. They law took effect, and prices fell afterward. argued that these characteristics On September 13, 1994, Title XI of the This suggests that the weapons became enhance offenders’ ability to kill and Federal Violent Crime Control and Law more available generally, but they must wound more persons and to inflict Enforcement Act of 1994—known as the have become less accessible to criminals multiple wounds on each victim, so Crime Control Act of 1994—took effect. because there was at least a short-term that a decrease in their use would Subtitle A (the Public Safety and Recre- decrease in criminal use of the banned reduce the fatality rate of gun ational Firearms Use Protection Act) of weapons. attacks. the act banned the manufacture, transfer, The ban’s impact on lethal gun and possession of certain semiautomatic Debated in a politically charged environ- violence is unclear because the firearms designated as assault weapons ment, the Public Safety and Recreational short period since the enabling and “large capacity” ammunition maga- Firearms Use Protection Act, as its title legislation’s passage created methodological difficulties for continued… R e s e a r c h i n B r i e f

Issues and Findings suggests, attempted to balance two or “sporterizing” a rifle by removing continued… competing policy goals. The first was to its pistol grip and replacing it with a respond to several mass shooting inci- thumbhole in the stock, for example, researchers. The National Institute dents committed with military-style and was sufficient to transform a banned of Justice is funding a followup study by the authors that is ex- other semiautomatics equipped with weapon into a legal substitute. On April pected to be released in 2000. It magazines holding large amounts of am- 5, 1998, President Clinton signed an will assess the longer term impacts munition. The second consideration was to Executive order banning the imports of of the ban and the effects of the limit the impact of the ban on recreational 58 foreign-made substitutes. other firearms provisions of Title XI. gun use by law-abiding owners, dealers, The long-term impacts of the ban and manufacturers. The ban specifically Gun bans and gun crime could differ substantially from the prohibited only nine narrow categories of short-term impacts. pistols, rifles, and shotguns (see exhibit 1). Evidence is mixed about the effectiveness It also banned “features test” weapons, that of previous gun bans. Federal restrictions Key findings: The authors, using enacted in 1934 on the ownership of fully a variety of national and local data is, semiautomatics with multiple features (e.g., detachable magazines, flash suppres- automatic weapons (machine guns) ap- sources, examined market trends— pear to have been quite successful based prices, production, and thefts—for sors, folding rifle stocks, and threaded bar- on the rarity with which such guns are the banned weapons and close rels for attaching silencers) that appeared 1 substitutes before estimating useful in military and criminal applications used in crime. Washington, D.C.’s re- potential ban effects and their but that were deemed unnecessary in strictive handgun licensing system, which consequences. shooting sports (see exhibit 2). The law also went into effect in 1976, produced a drop banned revolving cylinder shotguns (large in gun fatalities that lasted for several ● The research shows that the 2 capacity shotguns) and “large capacity years after its enactment. Yet, State and ban triggered speculative price in- magazines,” defined as ammunition- local bans on handguns have been found creases and ramped-up production to be ineffective in other research.3 of the banned firearms prior to the feeding devices designed to hold more law’s implementation, followed by than 10 rounds, far more than a hunter or The inconsistency of previous findings a substantial postban drop in prices competitive shooter might reasonably need may reflect, in part, the interplay of sev- to levels of previous years. (see exhibit 3). eral effects that a ban may have on gun ● Criminal use of the banned guns Various provisions of the ban limited markets. To reduce criminal use of guns declined, at least temporarily, after its potential effects on criminal use. As and the tragic consequences of such use, the law went into effect, which shown in exhibit 1, about half the banned a ban must make the existing stockpile suggests that the legal stock of makes and models were rifles, which are of guns less accessible to criminals (see preban assault weapons was, at hard to conceal for criminal use. Imports exhibit 4) by, for example, raising their 4 least for the short term, largely in of the five foreign rifle categories on this purchase prices. However, the anticipa- the hands of collectors and dealers. list had been banned in 1989. Further, tion of higher prices may encourage gun manufacturers to boost production just ● Evidence suggests that the ban the banned guns are used in only a small may have contributed to a reduc- fraction of gun crimes; even before the before the ban takes effect in the hope of tion in the gun murder rate and ban, most of them rarely turned up in generating large profits from the soon-to- murders of police officers by crimi- law enforcement agencies’ requests to the be collectors’ items. Immediately after the nals armed with assault weapons. Bureau of Alcohol, Tobacco and Firearms ban, criminals may find it difficult to pur- (BATF) to trace the sales histories of guns chase banned weapons if they remain in ● The ban has failed to reduce the recovered in criminal investigations. dealers’ and speculators’ storage facili- average number of victims per ties. Over the long term, however, the gun murder incident or multiple As a matter of equity, the law exempted stockpiled weapons might begin flowing gunshot wound victims. “grandfathered” guns and magazines into criminals’ hands, through straw pur- Target audience: Congressional manufactured before the ban took effect. chases, thefts, or “off-the-books” sales representatives and staff; State and While it also banned “exact” or duplicate that dealers or speculators falsely report local legislators; Federal, State, and copies of the prohibited makes and mod- to insurance companies and government local law enforcement officials; els, the emphasis was on “exact.” Short- officials as thefts.5 criminal justice practitioners and ening a gun’s barrel by a few millimeters researchers; advocacy groups; State and local government officials. 2 R e s e a r c h i n B r i e f

Exhibit 1. Description of firearms banned in Title XI

Name of Description 1993 Blue Preban 1993 BATF Examples of firearm Book price Federal legal trace request legal substitutes status status count

Avtomat Chinese, Russian, other foreign, and $550 (plus Imports 87 Norinco NHM* Kalashnikov domestic: 0.223 or 7.62x39mm caliber, 10–15% for banned in 90/91 (AK) semiautomatic Kalashnikov rifle, 5-, 10-, folding stock 1989 or 30-shot magazine, may be supplied models) with bayonet.

Uzi, Galil Israeli: 9mm, 0.41, or 0.45 caliber $550–$1,050 Imports 281 Uzi; Uzi Sporter** semiautomatic carbine, minicarbine, or (Uzi) banned in 12 Galil pistol. Magazine capacity of 16, 20, or 25, $875–$1,150 1989 depending on model and type (10 or (Galil) 20 on pistols).

Beretta Italian: 0.222 or 0.223 caliber, semiauto- $1,050 Imports 1 AR–70 matic paramilitary design rifle, 5-, 8-, banned in or 30-shot magazine. 1989

Colt AR–15 Domestic: Primarily 0.223 caliber $825–$1,325 Legal (civilian 581 Colt; Colt Sporter, paramilitary rifle or carbine, 5-shot version of 99 other Match H–Bar, magazine, often comes with two 5-shot military M–16) manufacturers Target; Olympic detachable magazines. Exact copies by PCR Models. DPMS, Eagle, Olympic, and others.

FN/FAL, Belgian design: 0.308 Winchester caliber, $1,100–$2,500 Imports 9 L1A1 Sporter** FN/LAR, FNC semiautomatic rifle or 0.223 Remington banned in (FN, Century) combat carbine with 30-shot magazine. 1989 Rifle comes with flash hider, 4-position fire selector on automatic models. Manufacturing discontinued in 1988.

SWD M–10 Domestic: 9mm paramilitary semiauto- $215 Legal 878 Cobray PM–11, M–11, matic pistol, fires from closed bolt, PM–12; Kimel M–11/9, 32-shot magazine. Also available in fully AP–9, Mini AP–9 M–12 automatic variation.

Steyr AUG Austrian: 0.223 Remington/5.56mm caliber, $2,500 Imports banned 4 semiautomatic paramilitary design rifle. in 1989

TEC–9 Domestic: 9mm semiautomatic paramilitary $145–$295 Legal 1202 Intratec; TEC–AB TEC DC–9, design pistol, 10- or 32-shot magazine; 175 Exact TEC–22 0.22 LR semiautomatic paramilitary copies design pistol, 30-shot magazine.

Revolving Domestic: 12 gauge, 12-shot rotary $525*** Legal 64 SWD Street Cylinder magazine, paramilitary configuration, Sweepers Shotguns double action.

* Imports were halted in 1994 under the Federal embargo on the importation of firearms from China. ** Imports banned by Federal Executive order, April 1998. *** Street Sweeper. Source for firearm descriptions: Blue Book of Gun Values, 17th ed., by S.P. Fjestad, 1996.

3 R e s e a r c h i n B r i e f

The timing and magnitudes of these Exhibit 2. Features test of the assault weapons ban market effects cannot be known in ad- vance. Therefore, the study examined market trends—prices, production, Flash Suppressor and thefts—for the banned weapons and close substitutes before estimating potential ban effects on their use and the consequences of that use. Barrel Mount Market effects Primary market prices of the banned guns and magazines rose by upwards of 50 percent during 1993 and 1994, Folding Stock while the ban was being debated in Congress. Gun distributors, dealers, High Capacity Detachable Magazine and collectors speculated that the banned weapons would become expen- sive collectors’ items. However, prices Pistol Grip fell sharply after the ban was implemented. Exhibit 4 shows price trends for a number of firearms. Prices for banned AR–15 rifles, exact copies, Exhibit provided courtesy of Handgun Control, Inc. and legal substitutes at least doubled in the year preceding the ban, fell to 1. Semiautomatic rifles having the ability to accept a detachable ammunition magazine near 1992 levels once the ban took and at least two of the following traits: effect, and remained at those levels ● A folding or telescoping stock. at least through mid-1996. Similarly, ● A pistol grip that protrudes beneath the firing action. prices of banned SWD semiautomatic ● A bayonet mount. pistols rose by about 47␣ percent during ● A flash hider or a threaded barrel designed to accommodate one. the year preceding the ban but fell by ● A grenade launcher. about 20␣ percent the following year. 2. Semiautomatic pistols having the ability to accept a detachable ammunition magazine For comparison, exhibit 4 shows that and at least two of the following traits: the prices of unbanned Davis and ● An ammunition magazine attaching outside the pistol grip. Lorcin semiautomatic pistols (among ● A threaded barrel capable of accepting a barrel extender, flash hider, forward the crime guns police seize most fre- handgrip, or silencer. quently) remained virtually constant ● A heat shroud attached to or encircling the barrel (this permits the shooter to 6 hold the firearm with the nontrigger hand without being burned). over the entire period. ● A weight of more than 50 ounces unloaded. Fueled by the preban speculative price ● A semiautomatic version of a fully automatic firearm. boom, production of assault weapons 3. Semiautomatic shotguns having at least two of the following traits: surged in the months leading up to the ● A folding or telescoping stock. ban. Data limitations preclude precise ● A pistol grip that protrudes beneath the firing action. and comprehensive counts. However, ● A fixed magazine capacity of more than five rounds. estimates based on BATF gun produc- ● Ability to accept a detachable ammunition magazine. tion data suggest that the annual production of five categories of assault weapons—AR–15s, models by Intratec, Note: A semiautomatic firearm discharges one shot for each pull of the trigger. After being fired, a semiautomatic cocks itself for refiring and loads a new round (i.e., bullet) automatically. SWD, AA Arms, and Calico—and legal substitutes rose by more than 120

4 R e s e a r c h i n B r i e f percent, from an estimated average of Exhibit 3. Logic model for Public Safety and Recreational Firearms Use 91,000 guns annually between 1989 Protection Act impact study and 1993 to about 204,000 in 1994— more than 1 year’s extra supply (see exhibit 5). In contrast, production of Title XI: Primary & AW/Magazine Consequences of unbanned Lorcin and Davis pistols fell Subtitle A Secondary Use in Crime Criminal Use Markets by about 35␣ percent, from an average of • Total • Gun murders • Price • Violent • Victims per event 283,000 annually between 1989 and • Production • Wounds per victim 1993 to 184,000 in 1994. • Thefts • Law enforcement officers killed These trends suggest that the preban price and production increases re- flected speculation that grandfathered weapons and magazines in the banned Exhibit 4. Comparison of price trends for banned and unbanned weapons categories would become profitable collectors’ items after the ban took effect. Instead, assault weapons prices 120 fell sharply within months after the 100 ban was in place, apparently under the combined weight of preban overpro- 80 duction of grandfathered guns and the introduction of new legal substitute 60 guns at that time.

These findings resemble what hap- 40 pened in 1989, when imports of sev-

Percentage of price at ban 20 eral models of assault rifles surged prior to the implementation of a Fed- 0 eral ban.7 Shortly thereafter, while California debated its own ban, crimi- 8 nal use of assault weapons declined, Jan–June ’92 July–Dec ’92 Jan–June ’93 July–Dec ’93 Jan–June ’94 July–Dec ’94 Jan–June ’95 July–Dec ’95 suggesting that higher prices and speculative stockpiling made the guns SWD handguns (B) Davis, Lorcin Semiauto handgun AR–15-type rifle less accessible to criminal users.9 Data were collected from display ads in randomly selected issues of the nationally distributed periodical It was plausible that the price and pro- Shotgun News. Price indices are adjusted for the mix of products and distributors advertised during duction trends related to the 1994 ban each time period. SWD, Davis, and Lorcin handgun data are reported semiannually. would be followed by an increase in reported thefts of assault weapons, for at sell the guns to ineligible purchasers mained largely in dealers’ and collec- least two reasons. First, if short-term and then falsely report them as stolen to tors’ inventories instead of leaking into price increases in primary markets tem- insurance companies and regulatory the secondary markets through which porarily kept assault weapons from en- agencies.10 criminals tend to obtain guns. tering illegal sales channels, criminals might be tempted to steal them instead. By the spring of 1996, however, there Criminal use of assault In addition, dealers and collectors had been no such increase. Instead, who paid high speculative prices for thefts of assault weapons declined weapons grandfathered assault weapons around about 14 percent as a fraction of all Because crime guns tend to be newly the time of the ban, but then watched as thefts of semiautomatics.11 Therefore, it purchased guns,12 it was hypothesized their investment depreciated after the appears that, at least in the short term, that speculative price increases would ban took effect, might be inclined to the grandfathered assault weapons re- tend to channel the flow of banned

5 R e s e a r c h i n B r i e f weapons from criminal purchasers to tion increase and the postban fall in available national sample of the types law-abiding speculators, thereby poten- prices. To estimate the net effect on of guns used in crime.13 These data are tially decreasing their use in criminal criminal use, the researchers measured limited because police agencies do not activities. (See “Study Design and criminal use of assault weapons using submit a trace request on every gun they Method.”) However, the potential de- data on gun trace requests submitted by confiscate. Many agencies submit very crease in criminal uses of the banned law enforcement agencies to BATF, few requests to BATF, particularly in weapons might be offset by the produc- whose tracing data provide the only States that maintain gun sales databases (such as California). Therefore, tracing data are a biased sample of guns recov- Exhibit 5. Production trends estimates for banned assault weapons and ered by police. Prior studies suggest that comparison guns* assault weapons are more likely to be (1) (2) (3) (4) submitted for tracing than are other 14 Firearm type 1994 1989–93 Ratio “Excess” confiscated firearms. production average [(1)/(2)] production production [(1)–(2)] As shown in exhibit 6, law enforcement agency requests for BATF as- AR–15 group 66,042 38,511 1.714 27,531 sault weapons traces in the 1993–95 Intratec 9mm, 22 102,682 33,578 3.058 69,104 period declined 20 percent in the first SWD family (all) and MAC (all) 14,380 10,508 1.368 3,872 calendar year after the ban took effect, AA Arms 17,280 6,561 2.633 10,719 dropping from 4,077 in 1994 to 3,268 Calico 9mm, 22 3,194 1,979 1.613 1,215 in 1995. Some of this decrease may Lorcin and Davis 184,139 282,603 0.652 reflect an overall decrease in gun Assault weapon total** 203,578 91,137 2.233 112,441 crimes; total trace requests dropped * Estimates are based on figures provided by gun manufacturers to BATF and compiled and 11␣ percent from 1994 to 1995, and disseminated annually by the Violence Policy Center. gun murders declined 10␣ percent over ** Assault weapon total excludes Lorcin/Davis group. the same period. Nevertheless, these trends suggest a 9- to 10-percent additional decrease (labeled with a triangle Exhibit 6. Relative changes in total and assault weapons traces in exhibit 6) due to substitution of other guns for the banned assault weapons in 1995 gun crimes.15 120 In contrast, assault weapons trace re- 110 quests from States with their own assault weapons bans declined by only an estimated 6 to 8 percent in 1995—further 100 evidence that the national trends reflect effects of the Federal ban. There were 90 fewer assault weapons traces in 1995 than in 1993 (3,748), suggesting that the 80 national decrease was not the result of a surge of assault weapons tracing around Percentage of 1994 level 16 70 the effective date of the ban. These national findings were sup- 60 1993 1994 1995 ported by analyses of trends in assault weapons recovered in crimes in St. Assault weapons traces Total gun traces Gun murders Louis and Boston, two cities that did not have preexisting State assault 1994=100 weapons bans in place. Although

6 R e s e a r c h i n B r i e f

Study design and method

S ubtitle A of Title XI banned the ● Comparison of gun production data measured nationally from Supplementary manufacture, transfer, and possession through 1994, the latest available year. Homicide Reports. of assault weapons and large capacity magazines. Researchers hypothesized ● Comparison and time series analyses ● Descriptive analysis of the use of that the ban would: of “leakage” of guns to illegal markets as assault weapons in mass murders in the measured by guns reported stolen to the United States from 1992 to 1996. ● Produce direct effects in the primary Federal Bureau of Investigation/National markets for these weapons. Crime Information Center between 1992 ● Comparison of data gathered between and 1996. 1992 and 1996 from medical examiners, ● Reduce, through related indirect one hospital emergency department, and effects in the secondary markets, the use The analysis of assault weapon use one police department in selected cities of these weapons in criminal activities. included: regarding the number of wounds per gunshot victim. ● Reduce the consequences of criminal ● Analysis of requests for BATF traces of gun use as measured by gun homicides assault weapons (1992–96) recovered in ● Analysis of 1992–96 data of law and, especially, incidents of multiple vic- crime investigations, both in absolute enforcement officers killed in action with tims, multiple wounds, and killings of law terms and as a percentage of all requests. assault weapons. enforcement officers. ● Preban and postban comparisons and For comparison purposes, researchers Because the measures of available data analyses of gun counts recovered in crime examined trends of types of guns and on these effects varied widely, the re- investigations by selected local law en- magazines that were affected differently search team decided to conduct several forcement agencies. by the ban. Few available databases re- small studies with different error sources late the consequences of assault weapon and integrate the findings. The strategy The analyses of the consequences of us- use to the make and model of the was to test whether the assault weapons ing assault weapons and semiautomatics weapon, so most of the analyses of con- and magazine bans interrupted these with large capacity magazines in criminal sequences are based on treatment and trends over time. Researchers employed activities included: comparison jurisdictions defined by the various types of time series and multiple legal environments in which the incident ● Examination of State time series data regression analyses, simple before-and- occurred. For instance, California, Con- on gun murders with controls for the po- after comparisons, and graphical displays. necticut, Hawaii, and New Jersey had tential influence of legal, demographic, banned assault weapons before 1994. and economic variables of criminological The analysis of market impacts included: Although interstate traffickers can cir- importance. cumvent State bans, researchers hypoth- ● Pricing trends in the primary markets esized that the existence of these for banned semiautomatic weapons, ● Comparisons and time series analyses State-level bans reduced the impact of comparable legal firearms, and large ca- of trends between 1980 and 1995 in the Federal ban in those respective pacity magazines using 1992–96 national victims per gun homicide incidents as jurisdictions. distributors’ price lists. assault weapons recoveries were rare Consequences of assault short period of time, thereby increasing in those cities both before and after weapons use the expected number of injuries and the ban, they declined 29 and 24 deaths per criminal use. The study ex- percent, respectively, as a share of all A central argument for special regula- amined trends in the following conse- gun recoveries during late 1995 and tion of assault weapons and large ca- quences of gun use: gun murders, into 1996. Because these cities’ trends pacity magazines is that they facilitate victims per gun homicide incident, reflect all guns recovered in crime, the rapid firing of high numbers of wounds per gunshot victim, and, to a they are not subject to the potential shots, which allows offenders to inflict lesser extent, gun murders of police. biases of trace request data. more wounds on more persons in a

7 R e s e a r c h i n B r i e f

the power of the statistical analyses to Exhibit 7. Estimated 1994–95 ban effects on total gun murder rate detect worthwhile ban effects that may have occurred. Given the limited use 42 usable of the banned guns and magazines in States gun crimes, even the maximum theo- -9.0%* retically achievable preventive effect of the ban on outcomes such as the gun murder rate is almost certainly State too small to detect statistically be- Yesassault No weapons cause the congressionally mandated ban? timeframe for the study effectively limited postban data collection to, at 4 States 38 States -0.1% -10.3%* most, 24 months (and only 1 calendar year for annual data series). Nevertheless, to estimate the first-year State State ban effect on gun murders, the analysis Yes juvenile No Yesjuvenile No compared actual 1995 State gun possession possession murder rates with the rates that would ban? ban? have been expected in the absence of the assault weapons ban. Data from 1980 to 1995 of 42 States with ad- 2 States 2 States 16 States 22 States equate annual murder statistics (as +4.5% -7.3% -10.9% -9.7% reported to the Federal Bureau of In- vestigation) were used to project 1995 gun murder rates adjusted for ongoing trends and demographic and economic changes. Tests were run to determine Drop California and New York whether the deviation from the projection could be explained by various policy interventions other than the assault weapons ban.

1 State 2 States 15 States 22 States Exhibit 7 displays the steps in that +5.8% -7.6% -6.7% -9.8% analysis. Overall, 1995 gun murder rates were 9 percent lower than the projection.19 Gun murders declined * Statistically significant at 10-percent level. 10.3 percent in States without preexisting assault weapons bans, but they remained unchanged in States with There were several reasons to expect, magazines are used in 20 to 25 per- their own bans. After adjusting the at best, a modest ban effect on crimi- cent of these gun crimes, it is not clear projection for possible effects of State nal gun injuries and deaths. First, how often large capacity magazines bans on juvenile handgun possession studies before the ban generally found actually turn a gun attack into a gun and a similar Federal ban that took that between less than 1 and 8 percent murder.18 Second, offenders could effect simultaneously with the assault of gun crimes involved assault weap- replace the banned guns with legal weapons ban, the study found that ons, depending on the specific defini- substitutes or other unbanned semiau- 1995 gun murder rates were 10.9 per- tion and data source used.17 Although tomatic weapons to commit their cent below the projected level. Finally, limited evidence suggests that semiau- crimes. Third, the schedule for this statistical controls were added for tomatics equipped with large capacity study set out in the legislation limited

8 R e s e a r c h i n B r i e f postban drops in California and New ficiently rare that the available data do the long-term consequences may York to avoid confounding possible not permit a reliable assessment of differ substantially from the short- effects of the assault weapons ban, whether this contributed to a general term consequences reported here. California’s “three strikes” law, and reduction in gun murders of police. (A followup study of longer term New York City’s “quality of life” polic- impacts of the ban and the effects ing. Still, 1995 murder rates in the 15 Implications and research of other provisions of Title XI is remaining States with juvenile hand- recommendations underway and is expected to be gun possession bans but no assault released in 2000.) weapons ban were 6.7 percent below It appears that the assault weapons ban ● Develop new gun market data the projection—a difference that had clear short-term effects on the gun sources and improve existing could not be explained in terms of market, some of which were unintended ones. For example, NIJ and BATF murder trends, demographic and eco- consequences: production of the should consider cooperating to nomic changes, the Federal juvenile banned weapons increased before the establish and maintain time series handgun possession ban, or the law took effect and prices fell after- data on primary and secondary mar- California and New York initiatives. ward. These effects suggest that the weapons became more available gener- ket prices and production of assault Random, year-to-year fluctuations ally, but they must have become less weapons, legal substitutes, other could not be ruled out as an explana- accessible to criminals because there guns commonly used in crime, and tion of the 6.7-percent drop. With only was at least a short-term decrease in the respective large and small ca- 1 year of postban data available and criminal use of the banned weapons. pacity magazines. Like similar sta- only 15 States meeting the screening Evidently, the excess stock of grand- tistical series currently maintained criteria for the final estimate, the model fathered assault weapons manufactured for illegal drugs, such a price and lacks the statistical power to detect a prior to the ban is, at least for now, production series would be a valu- preventive effect of even 20 percent un- largely in the hands of dealers and col- able instrument for monitoring der conventional standards of statistical lectors. The ban’s short-term impact on effects of policy changes and other reliability.20 Although it is highly im- gun violence has been uncertain, due influences on markets for weapons probable that the assault weapons ban perhaps to the continuing availability of that are commonly used in crime. produced an effect this large, the ban grandfathered assault weapons, close ● Examine potential substitution could have reduced murders by an substitute guns and large capacity effects. A key remaining question amount that would escape statistical magazines, and the relative rarity with is whether offenders who preferred detection. which the banned weapons were used the banned assault weapons have in gun violence even before the ban. switched to the new legal substitute However, other analyses using a variety of models or to other legal guns, such national and local data sources found no To provide a more current and detailed as semiautomatic handguns that clear ban effects on certain types of mur- understanding of the assault weapons accept large capacity magazines. ders that were thought to be more closely ban and gun markets generally, we associated with the rapid-fire features of recommend a variety of further steps: ● Study criminal use of large assault weapons and other semiautomat- capacity magazines. The lack of ics equipped with large capacity maga- ● Update the impact analysis. This knowledge about trends in the crimi- zines. The ban did not produce declines study was conducted with data col- nal use of large capacity magazines in the average number of victims per inci- lected within 24 months of the ban’s is especially salient for three rea- dent of gun murder or gun murder victims passage; a number of the analyses sons. The large capacity magazine is with multiple wounds. were conducted with only 1 calen- perhaps the most functionally impor- dar year of postban data. This lim- tant distinguishing feature of assault Murders of police by offenders armed ited timeframe weakens the ability weapons. The magazine ban also with assault weapons declined from an of statistical tests␣ to discern im- affected more gun models and gun estimated 16 percent of gun murders of pacts that may be meaningful from crimes than did the bans on desig- police in 1994 and early 1995 to 0 per- a policy perspective. Also, because nated firearms. Finally, recent cent in the latter half of 1995 and early the ban’s effects on gun markets anecdotal evidence suggests that 1996. However, such incidents are suf- and gun violence are still unfolding, new and remanufactured preban,

9 R e s e a r c h i n B r i e f

high-capacity magazines are begin- simple comparisons; much more are still being refined, and none will ever ning to reappear in the market for comprehensive research that takes stop all gun violence. However, with dis- use with legal semiautomatic pistols. into account important characteris- passionate analyses of their effects and a tics of the actors and situations willingness to modify tactics in response ● Improve the recording of should be pursued. Future research to evidence, these approaches may well magazines recovered with crime on the dynamics of criminal prove more immediately effective, and guns. To better understand the role shootings, including various mea- certainly less controversial, than large capacity magazines play in gun sures of the number of shots fired, regulatory approaches alone. crimes, BATF and State and local wounds inflicted, and victims killed law enforcement agencies should or wounded, would improve esti- encourage efforts to record the Notes mates of the potential effects of the magazines with which confiscated 1. Kleck, Gary, Point Blank: Guns and assault weapons and magazine ban, firearms are equipped—information Violence in America, New York: Aldine De while yielding useful information on that frequently goes unrecorded un- Gruyter, 1991. violent gun crime generally. der current practice. Further studies 2. Loftin, Colin, David McDowall, Brian are needed on trends in the criminal Future directions Wiersema, and Talbert J. Cottey, “Effects of use of guns equipped with large Restrictive Licensing of Handguns on Homicide capacity magazines. Gun control policies, and especially and Suicide in the District of Columbia,” New gun bans, are highly controversial England Journal of Medicine, 325: 1625–1630. ● Conduct indepth, incident- crime control measures, and the debates based research on the situ- 3. Kleck, Point Blank: Guns and Violence in tend to be dominated by anecdotes and America. ational dynamics of fatal and emotion rather than empirical findings. nonfatal gun assaults. Despite 4. The ban exempted assault weapons manu- In the course of this study, the research- the rhetoric that characterizes fire- factured before the effective date of the law. ers attempted to develop a logical arms policy debates, there are still Because significant deterioration or loss of framework for evaluating gun policies, those guns occurs only over decades, any im- questions regarding the impacts one that considers the workings of gun mediate ban effects would have to reflect scar- that weaponry, actor, and situ- markets and the variety of outcomes city of assault weapons to criminal purchasers, ational characteristics have on the rather than a dwindling of the stockpile. such policies may have. The findings outcomes of gun attacks. Therefore, suggest that the relatively modest gun 5. A number of researchers and journalists research is needed to gain a greater control measures that are politically have commented on the weak state of Federal understanding of the roles of feasible in this country may affect gun firearms licensees (FFLs) regulation, particu- banned and other weapons in inten- larly before 1994 when Title XI strengthened markets in ways that at least temporarily tional deaths and injuries. In what the screening process for obtaining and renew- reduce criminals’ access to the regu- percentage of gun attacks, for in- ing licenses. Empirical evidence suggests that lated guns, with little impact on law- a small minority of gun dealers supply many of stance, does the ability to fire more abiding gun owners. the guns used by criminals. Analysis of Bureau than 10 rounds without reloading of Alcohol, Tobacco and Firearms tracing data influence the number of gunshot The public safety benefits of the 1994 by Glenn Pierce and his colleagues in 1995 wound victims or determine the dif- ban have not yet been demonstrated. showed that while 92 percent of FFLs had no ference between a fatal and nonfatal confiscated guns traced back to them, 0.4 per- This suggests that existing regulations cent of the dealers were linked to nearly 50 attack? The study yielded some should be complemented by further tests percent of the traced weapons. Although some weak evidence that victims killed of enforcement tactics that focus on the of this concentration could simply reflect the by guns having large capacity tiny minorities of gun dealers and owners proximity of some large law-abiding dealers to magazines (including assault weap- who are linked to gun violence. These in- high-crime areas, evidence suggests that illegal ons) tend to suffer more bullet practices by some dealers contribute to this clude strategic targeting of problem gun concentration. See Wachtel, Julius, “Sources of wounds than victims killed with dealers,21 crackdowns on “hot spots” for Crime Guns in Los Angeles, California,” Polic- other firearms and that mass mur- gun crime,22 and strategic crackdowns on ing: An International Journal of Police Strate- ders with assault weapons tend to perpetrators of gun violence,23 followed gies and Management, 21(2) (1998): 220–239; involve more victims than those by comprehensive efforts to involve com- Larson, Erik, Lethal Passage: The Story of a with other firearms. However, Gun, New York: Vintage Books, 1995; Pierce, munities in maintaining the safety that Glenn L., LeBaron Briggs, and David A. research results were based on these tactics achieve.24 These techniques

10 R e s e a r c h i n B r i e f

Carlson, The Identification of Patterns in Fire- Gun transfers made by nonlicensed citizens do U.S. Department of Justice, Bureau of Justice arms Trafficking: Implications for Focused not require such recordkeeping. In some in- Statistics, 1993; Hargarten, Stephen W., Trudy Enforcement Strategy, Washington, D.C.: U.S. stances, however, gun owners who knowingly A. Karlson, Mallory O’Brien, Jerry Hancock, Department of the Treasury, Bureau of Alcohol, transfer guns to ineligible purchasers may and Edward Quebbeman, “Characteristics of Tobacco and Firearms, 1995; and Violence choose to falsely report the guns as stolen to Firearms Involved in Fatalities,” Journal of the Policy Center, More Gun Dealers Than Gas Sta- prevent themselves from being linked to any American Medical Association, 275 (1996): tions: A Study of Federally Licensed Firearms crimes committed with the guns. 42–45; Hutson, H. Range, Deirdre Anglin, and Dealers in America, Washington, D.C.: Michael J. Pratts, Jr., “Adolescents and Chil- Violence Policy Center, 1992. 11. This finding is a revision of results reported dren Injured or Killed in Drive-by Shootings in Chapter 4 of Roth and Koper, Impact Evalu- in Los Angeles,” The New England Journal of 6. Like assault weapons prices, large capacity ation of the Public Safety and Recreational Medicine, 330 (1994): 324–327; Kleck, Gary, magazine prices generally doubled in the year Firearms Use Protection Act of 1994. Targeting Guns: Firearms and Their Control, preceding the ban. However, trends diverged New York: Aldine De Gruyter 1997; Cox News- after the ban, depending on the gun for which 12. Zimring, Franklin E., “Street Crime and New papers, Firepower: Assault Weapons in America, the magazine was made. See Chapter 4 in Roth, Guns: Some Implications for Firearms Control,” Washington, D.C.: Cox Newspapers, 1989; Jeffrey A., and Christopher S. Koper, Impact Journal of Criminal Justice, 4 (1976): 95–107. McGonigal, Michael D., John Cole, C. William Evaluation of the Public Safety and Recre- Schwab, Donald R. Kauder, Michael F. 13. A gun trace usually tracks a gun to its ational Firearms Use Protection Act of 1994, Rotondo, and Peter B. Angood, “Urban Firearm first point of sale by a licensed dealer. Upon Washington, D.C.: The Urban Institute, 1997. Deaths: A Five-Year Perspective,” The Journal request, BATF traces guns suspected of being of Trauma, 35 (1993): 532–536; New York used in crime as a service to Federal, State, and 7. American Medical Association Council on State Division of Criminal Justice Services, local law enforcement agencies. Scientific Affairs, “Assault Weapons as a Assault Weapons and Homicide in New York Public Health Hazard in the United States,” 14. For additional discussions of the limits of City, Albany, New York: New York State Divi- Journal of the American Medical Association, sion of Criminal Justice Services, 1994; Zawitz, 267 (1992): 3067–3070. tracing data, see Chapter 5 in Roth and Koper, Impact Evaluation of the Public Safety and Rec- Marianne W., Guns Used in Crime; also see review in Koper, Christopher S., Gun Lethality 8. Mathews, J., “Unholstering the Gun Ban,” reational Firearms Use Protection Act of 1994; and Homicide: Gun Types Used by Criminals The Washington Post, December 31, 1989. Zawitz, Marianne W., Guns Used in Crime, Washington, D.C.: U.S. Department of Justice, and the Lethality of Gun Violence in Kansas 9. Cook, Philip J., and James A. Leitzel, Bureau of Justice Statistics, 1995; and Kleck, City, Missouri, 1985–1993, Ann Arbor, “ ‘Perversity, Futility, Jeopardy’: An Economic Gary, Targeting Guns: Firearms and Their Michigan: University Microfilms, Inc., 1995. Analysis of the Attack on Gun Control,” Law Control, New York: Aldine De Gruyter, 1997. 18. See Chapter 6 in Roth and Koper, Impact and Contemporary Problems, 59 (1996): Evaluation of the Public Safety and Recre- 91–118. 15. Percentage decreases in assault weapon traces related to violent and drug crimes were ational Firearms Use Protection Act of 1994; 10. Since enactment of the Gun Control Act of similar to or greater than those for total assault and New York State Division of Criminal Jus- 1968, FFLs are required to retain records of all weapons, although these categories were quite tice Services, Assault Weapons and Homicide in gun sales and a running log of their gun acqui- small in number. Separate analyses were con- New York City. sitions and dispositions. Federal law has ducted for all assault weapons and for a select 19. In addition to the variables discussed in the various regulations governing sales by FFLs, group of domestically produced assault weap- text, the models included an indicator variable including the requirement that FFLs have po- ons that were still in production when the ban for each State, a polynomial time trend for the tential gun purchasers sign statements that they went into effect. Both analyses showed the same national gun homicide trend, and annual State- are not legally ineligible to purchase firearms. drop in overall trace requests. See Chapter 5 in level controls for per capita income, employ- The 1993 Brady Act further requires FFLs to Roth and Koper, Impact Evaluation of the Pub- ment rates, and age structure of the population. obtain photo identification of potential handgun lic Safety and Recreational Firearms Use purchasers, notify the chief local law enforce- Protection Act of 1994. 20. By conventional standards, we mean statis- ment officer of each application for a handgun tical power of 0.8 to detect a change, with 0.05 16. In general, our analysis of assault weapons purchase, and wait 5 business days before com- probability of a Type I error. pleting the sale, during which time the chief use did not include legal substitute versions of law enforcement officer may check the the banned weapons. However, lack of preci- 21. Pierce et al., The Identification of Patterns applicant’s eligibility. sion in the data sources could have resulted in Firearms Trafficking: Implications for in some of these weapons being counted as Focused Enforcement Strategy. FFLs who sell guns without following these re- postban traces or recoveries of assault weapons. quirements may, if inspected by BATF, try to 22. Sherman, Lawrence W., James W. Shaw, cover up their illegal sales by claiming that the 17. For example, see Beck, Allen, Darrell and Dennis P. Rogan, The Kansas City Gun Ex- guns were lost or stolen. To help prevent such Gilliard, Lawrence Greenfeld, Caroline Harlow, periment, Research in Brief, Washington, D.C.: practices, Subtitle C of Title XI requires FFLs Thomas Hester, Louis Jankowski, Tracy Snell, U.S. Department of Justice, National Institute to report all stolen and lost firearms to BATF James Stephan, and Danielle Morton, Survey of of Justice, 1995, NCJ 150855 and local authorities within 48 hours. State Prison Inmates, 1991, Washington, D.C.:

11 R e s e a r c h i n B r i e f

23. Kennedy, David, Anne M. Piehl, and 24. Kelling, G.L., M.R. Hochberg, S.K. Anthony A. Braga, “Youth Violence in Boston: Costello, A.M. Rocheleau, D.P. Rosenbaum, The National Institute of Justice is a Gun Markets, Serious Youth Offenders, and J.A. Roth, and W.G. Skogan, The Bureau of component of the Office of Justice a Use-Reduction Strategy,” Law and Justice Assistance Comprehensive Communities Programs, which also includes the Bureau Contemporary Problems, 59 (1996): 147–196. Program: A Preliminary Report, Cambridge, of Justice Assistance, the Bureau of Justice Massachusetts: Botec Analysis (forthcoming). Statistics, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. Jeffrey A. Roth, Ph.D., is a principal research associate at the State Policy Center of The Urban Institute, and Christopher S. Koper, Ph.D., is a research Findings and conclusions of the research associate at the State Policy Center of The Urban Institute. The views ex- reported here are those of the authors and do pressed are those of the authors and should not be attributed to the Urban In- not necessarily reflect the official position or stitute, its trustees, or its funders. The research for this study was supported by policies of the U.S. Department of Justice. NIJ grant 95–IJ–CX–0111. NCJ 173405

This and other NIJ publications can be found at and downloaded from the NIJ Web site (http://www.ojp.usdoj.gov/nij).

U.S. Department of Justice PRESORTED STANDARD Office of Justice Programs POSTAGE & FEES PAID National Institute of Justice DOJ/NIJ PERMIT NO. G–91 Washington, DC 20531

Official Business Penalty for Private Use $300