Criminal Justice Interventions to Reduce Firearm-Related Violence
This chapter reviews the state of knowledge of the effectiveness of criminal justice interventions aimed at reducing deliberate or accidental injuries or deaths from firearms. The policies are: (1) gun courts, (2) enhanced sentences for criminal uses of firearms, and (3) problem-oriented policing to prevent firearm-related crimes. These interventions have had recent broad bipartisan support and are a major focus of the federal government’s ongoing efforts to reduce firearm-related violence. In particular, over $500 million has been devoted to Project Safe Neighborhoods, a program designed to provide funds to hire new federal and state prosecutors, support investigators, provide training, and develop and promote community outreach efforts (for further details, see http://www.psn.gov/about.asp\). The research evidence, however, is mixed. In some cases, the committee found evidence that programs may be effective, in others the evidence suggests that programs may have negligible effects, and in others the evidence base is lacking.
Gun courts, which are descendants of the drug court movement of the 1990s, generally target particular types of offenders for quicker, and sometimes tougher, processing in community-based courts. Gun courts operate differently across jurisdictions but typically feature small caseloads, frequent hearings, immediate sanctions, family involvement, and treatment services. Little research has been conducted on the operations and crime prevention effectiveness of gun courts. Most available knowledge comes
from the Office of Juvenile Justice and Delinquency Prevention’s examination of a juvenile gun court operating in Jefferson County, Alabama.
The Jefferson County Juvenile Gun Court in Birmingham, Alabama, focuses on first-time juvenile gun offenders. Its core components include a 28-day boot camp, a parent education program, a substance abuse program, intensive follow-up supervision, and community service (Office of Juvenile Justice and Delinquency Prevention, 2002). Birmingham’s juvenile gun court is administered as part of the family court and provides services to offenders and their families. The juvenile gun court seeks to provide swift consequences by reviewing incoming cases within 72 hours and trying them within 10 working days. The court also attempts to provide certain consequences by providing judges with the authority to impose mandatory detention of juvenile offenders, with judicial discretion as to whether juvenile cases are eligible for diversion. All offenders attend the 28-day boot camp, and the court can add more time to a youth’s stay for various infractions. While the juveniles attend boot camp, parents attend an education program that includes training on improving youth-parent communication skills and discussions of the impact of firearm-related violence on victims, perpetrators, and families. Parents who fail to complete the program may be arrested and jailed. After the youths return from boot camp, they are required to participate in substance abuse classes for six weeks, take mandatory weekly drug tests during this time period, and perform community service work, such as neighborhood and graffiti cleanup. Probation officers and transition aides provide intensive follow-up supervision, and parental involvement is required throughout the adjudication process.
An evaluation of the Birmingham juvenile gun court compared the case processing records and recidivism rates for three groups of juvenile gun offenders: a group of Birmingham juveniles with limited prior offenses who participated in the gun court’s core components, a group of Birmingham juveniles with prior offenses who received short juvenile correction commitments and did not receive after-care monitoring, and a comparison group of juveniles from a nearby city who did not participate in a gun court program (Office of Juvenile Justice and Delinquency Prevention, 2002). The evaluation revealed that the Birmingham gun court group had significantly lower levels of recidivism (17 percent) than the Birmingham nongun court group (37 percent) and the comparison group (40 percent). The evaluators also found that having a prior gun offense (common to youth in the nongun court groups) increased the odds of recidivism (Office of Juvenile Justice and Delinquency Prevention, 2002). The evaluation did not provide an estimate of the extent to which the differences among the groups in prior gun offending could account for some of the observed recidivism reductions.
ENHANCED SENTENCES FOR CRIMINAL USE OF FIREARMS
Sentencing Enhancements for Firearm-Related Crimes
Firearms sentence enhancement laws mandate minimum sentences or extra prison time for felonies committed with firearms. Unlike most gun control measures, enhanced prison penalties for firearm-related crimes have widespread support from all sides of the firearms policy debate. Firearms sentence enhancements do not affect the ability of law-abiding citizens to keep firearms for recreation or self-defense and have the potential to reduce firearm-related violence by incapacitating those who have been convicted of firearm-related crimes and deterring future firearms crimes. Although a recent rigorous research study suggests that sentence enhancements can result in modest crime reductions (Kessler and Levitt, 1999), the evidence on the effects of sentencing enhancements on firearm-related crime is less clear.1
In their examination of the case for a gun-emphasis policy in the prosecution of violent offenders, Cook and Nagin (1979) conclude that firearms use in robbery and assault deserves stiffer punishment because it increases the chance of the victim’s death. In their analysis of case information, Cook and Nagin found that, while there was little difference in recidivism rates between gun users and those using other weapons in Washington, DC, criminals who used a gun in one crime were more likely to be rearrested for a firearm-related crime. Finally, they also found that, in Washington during the mid-1970s, there was little distinction in prosecution and sentencing between firearm-related crime defendants and other-weapon-related defendants when controlling for other characteristics of the case. Apparently, prosecutors had a “weapons” emphasis, but not a “gun” emphasis.
Several small-scale studies suggest that sentencing enhancements for firearm-related crimes might reduce some types of crimes. The results of these studies, however, are sometimes difficult to interpret. For example, McPheters and his colleagues (1984) used interrupted-time-series analyses to evaluate the effects of Arizona’s 1974 firearms sentencing enhancement law. They found highly significant reductions in firearm-related robberies in Pima and Maricopa counties and no significant firearm-related robbery reductions in five southwestern cities outside Arizona that did not pass similar laws during the study time period used as controls. This impact on firearm-related robberies, however, may have been due to regression to the mean, as Arizona experienced a 75 percent increase in firearm-related robberies in the two years prior to the passage of the law (McPheters et al., 1984).
Loftin and McDowall (1981) examined the effects of a Michigan firearms law that required a 2-year mandatory sentence for felonies committed while in possession of a firearm on the certainty and severity of sentences and on the number of serious violent crimes in Detroit. A substantial media campaign announcing that “One With a Gun Gets You Two” preceded the law’s going into effect in January 1977, and the Detroit prosecutor adopted a strict policy of not plea bargaining such cases down to lesser charges. Their examination of cases processed though the Detroit Recorder’s Court between 1976 and 1978 found little change in the certainty or severity of sentences for firearm-related murders and armed robberies, but they did find that there was a significant increase in the expected sentence for firearms-related assault cases that could be attributed to the firearms law.
Similarly, in their examination of California’s firearms sentencing enhancement law, Lizotte and Zatz (1986) found little difference in prison sentences given to firearm-related criminals in California, except when defendants had three or more prior arrests. Loftin and McDowall (1981) evaluated the crime control effects of the Michigan firearms law, again using time-series analysis, and found no significant reductions in armed robbery or firearm-related assaults in Detroit. They did find a significant reduction in firearm-related homicides but concluded that the overall results best fit a model in which the firearms law had no preventive effects on crime. A later analysis by Loftin et al. (1983) affirmed these conclusions.
The Florida Felony Firearm Law mandated a 3-year prison sentence for anyone possessing a firearm while committing or attempting to commit any of 11 specified felonies. Using time-series analysis models, Loftin and McDowall (1984) examined the effects of the October 1975 Florida firearms law on firearm-related homicides, armed robberies, and firearm-related assaults in Miami, Tampa, and Jacksonville. To reduce any historical effects, nonfirearm-related homicides, unarmed robberies, and knife assaults were used as control time series. Loftin and McDowall (1984) did not find any significant reductions in firearm-related crime in Jacksonville and Miami associated with the passage of the gun law. They did, however, find a significant decrease in firearm-related homicides and a significant increase in firearm-related assaults in Tampa. While they recommend further testing and examination of the data, Loftin and McDowall (1984) tentatively concluded that the Florida firearms law did not have a measurable deterrent effect on violent crime. In a later paper, McDowall et al. (1992) pooled the Detroit, Jacksonville, Tampa, and Miami time series with data collected from a study examining the effects of Pennsylvania’s 1982 firearm sentencing enhancement law on violent crime in Pittsburgh and Philadelphia. They found that these two cities showed significant reductions in homicide associated with the passage of the law. The pooled results led McDowall and his colleagues (1992) to very different conclu-
sions from the city-level studies. The authors found that mandatory sentencing laws significantly reduced the number of homicides, but the effects of mandatory sentencing laws on assaults and robberies were inconclusive.
Nationwide studies have not found any crime prevention effects associated with firearms sentence enhancements. Kleck (1991) conducted a cross-sectional analysis of 1980 data for 170 cities and found that the existence of a firearms sentence enhancement law was not related to homicide, assault, or robbery rates. However, as Marvell and Moody (1995) observe, cross-sectional designs are not suitable for studying short-term impacts, and it is difficult to be confident that the control variables account for the numerous differences between cities that may mask the laws’ impacts. In an attempt to mitigate the methodological problems in earlier research studies, Marvell and Moody (1995) conducted a comprehensive evaluation of the effects of firearms sentence enhancements on crime and prisons. The authors conducted a pooled time-series, cross-sectional design analysis for nearly all states over a period of 16 to 24 years such that, for each state, the other states served as controls. They found little evidence to suggest that firearms sentencing enhancements had any effects on crime rates or firearms use. Moreover, the authors did not find any indication that these laws increased prison admissions or prison populations.
Raphael and Ludwig (2003) observe that Richmond’s well-known Project Exile program to deter illicit carrying of firearms by convicted felons is essentially a firearms sentence enhancement initiative, as firearms offenders are diverted from state to federal courts. At the heart of Project Exile, all Richmond felon-in-possession cases are prosecuted in federal courts, with the defendants facing a mandatory five-year prison sentence if convicted. Project Exile also includes training for local law enforcement on federal statutes and search and seizure procedures, a public relations campaign to increase community involvement in fighting firearm-related crime, and a massive advertising campaign intended to send the message of zero tolerance for gun crime and to inform potential offenders of the swift and certain federal sentence (Raphael and Ludwig, 2003). Advocates of the program claim success based on a 40 percent decrease in Richmond firearm-related homicides between 1997 and 1998.
In their evaluation of Project Exile, Raphael and Ludwig (2003) found that the decline in Richmond firearm-related homicides would have been likely to occur even in the absence of the program. The authors revealed that nearly all of the reduction in Richmond firearm-related homicides associated with Project Exile may be attributable to an unusually high level of and increase in firearm-related homicide prior to the implementation of the program. They also found little statistical evidence of an impact between felon-in-possession convictions and city-level homicide rates. Their null finding is robust to a variety of methodological adjustments, including
an analysis of omitted variable bias that uses juveniles, who are generally exempt from federal felon-in-possession charges, as an additional in-city control group. In his subsequent analysis of the Raphael-Ludwig data, Levitt (2003) suggests that the expectations of a large decrease in crime associated with Project Exile were probably unrealistic, given the small number of additional felon-in-possession convictions per year (roughly 80) and the small increase in total punishment in Richmond (240 extra person-years of imprisonment that would be associated with an estimated 2.5 percent reduction in crime in Richmond). Greenwood (2003) also speculates that Project Exile did not focus sufficiently on the most dangerous offenders associated with the bulk of firearm-related crime in Richmond.
Mandatory Penalties for Unlawful Carrying of Guns
Mandatory sentencing laws, which require a mandatory penalty for unlicensed or otherwise unlawful carrying of a firearm, seek to reduce gun use in unpremeditated crimes by deterring the casual carrying of firearms in public places. The best known example of laws that institute a mandatory penalty for unlawful carrying is the Massachusetts Bartley-Fox gun law.
The Massachusetts legislature enacted the Bartley-Fox gun law, which mandated a one-year minimum prison term for the unlicensed carrying of firearms and a two-year sentence for crimes committed while possessing a gun, to reduce the incidence of firearm-related crime as well as the illicit carrying of firearms (Beha, 1977). The amendment was adopted in July 1974 and became effective beginning in April 1975. Two months prior to the law’s effective date, a concerted campaign was launched to characterize the impending consequences in the following terms, “If you are caught with a gun, you will go to prison for a year and nobody can get you out” (Pierce and Bowers, 1981:122).
While the mandatory sentence provision removed most judicial discretion in sentencing a defendant convicted of illegally carrying a gun, the defendant could in fact escape the 1-year sentence in a variety of ways (Deutsch and Alt, 1977). If someone was apprehended with a firearm on his person, the police could file a charge of illegal possession, which does not carry a mandatory minimum, rather than a charge of illegal carrying. Later in the process, prosecutors could also press for the lesser possession charge regardless of the initial police charge. Judges and juries could also find the defendant guilty of a lesser charge. As Zimring commented, “the one-year minimum will only invoke mandatory one-year jail terms for carrying without a license to the extent that police, prosecutors, and judges want it to produce such results. If there is strong resistance from any single link in this chain, the mandatory minimum can be avoided” (as quoted in Deutsch and Alt, 1977:545).
A series of studies examined the impact of the Bartley-Fox law on gun crime and the administration of justice in Boston. Beha (1977) examined the daily application of the Bartley-Fox law by police, prosecutors, and judges through simple before-after analyses of prosecutions for firearms violations and firearm-related crimes in Boston between 1974 and 1975. He analyzed two 6-month samples of all complaints relating to the illegal use, possession, or carrying of a firearm for 1974 and 1975 drawn from the dockets of Boston district courts and cross-checked against Boston Police Department records.
His analysis suggests that Bartley-Fox made it more likely for a prison sentence to be imposed in both firearm assault prosecutions and cases in which illegal carrying was the most serious charge, but the law did not affect the disposition of prosecutions for armed robbery and homicide. Beha (1977) also found that criminal justice officials did not systematically attempt to evade the mandatory sentence. In an analysis of yearly issuances of firearms identification cards and licenses to carry firearms between 1970 and 1975, Beha reported that the high degree of publicity attendant on the amendment’s passage, some of which was inaccurate, increased citizen compliance with existing legal stipulations surrounding firearm acquisition and possession, some of which were not in fact addressed by the amendment. Using simple before-after analyses of percentage changes in reported crime rates between 1970 and early 1977, Beha notes that the law did not seem to affect armed robbery but produced definite reductions in firearm-related assaults and firearm-related assault-homicides. However, the total number of aggravated assaults remained constant over time, suggesting a shift from guns to less lethal weapons.
Other studies suggest that criminal justice practitioners may have hindered the implementation of the Bartley-Fox law. In interviews with Boston police officers, Carlson (1982) found that 89 percent of the officers interviewed reported becoming more selective about whom to frisk for weapons, as they did not want to arrest someone who was otherwise a law-abiding citizen. The National Institute of Justice also reports that, between 1974 and 1976, arrests in gun incidents decreased by 23 percent, while weapons seizures without arrest increased by 120 percent. While it is unclear whether the number of guns seized without arrest increased in tandem with all weapons seizures, these figures suggest that police may have made fewer gun-carrying arrests to evade the law.
Rossman et al. (1980) found that Bartley-Fox impeded the flow of cases through the criminal justice system, as defendants had no incentive to plead guilty. They found that the rate at which gun-carrying cases went to trial tripled, the conviction rate was halved, and the median time to disposition doubled. Dismissals and not-guilty verdicts doubled, suggesting that judges may have been avoiding the imposition of the mandatory sentence
for some defendants. Rossman and his colleagues (1980) also found that, while a smaller fraction of gun-carrying defendants were convicted of felony gun-carrying, the fraction that received prison sentences did increase. These shifts in case processing and the discretionary actions of criminal justice practitioners in Massachusetts are common responses to the adoption of mandatory sentences (see, e.g., Alschuler, 1978).
Pierce and Bowers (1981) used interrupted-time-series techniques and multiple control group comparisons to examine the impact of Bartley-Fox on firearm-related and nonfirearm-related assaults, robbery, and homicide in Boston. They found a statistically significant reduction in gun assaults in March 1975, one month prior to the implementation of the Bartley-Fox law. The authors suggest that the vigorous publicity campaign influenced behavior before the law actually went into effect. The multiple control group comparison consisted of simple percentage change analyses of firearm-related crime rates in 1974 and 1975 for Boston relative to other New England cities, the United States without Massachusetts, the middle Atlantic states, the north central states, and selected cities within a 750-mile radius, including Washington, DC, Baltimore, New York, Philadelphia, Cleveland, and Detroit. Pierce and Bowers (1981) found that the law significantly reduced firearm-related assaults, but produced offsetting increases in nonfirearm-related armed assaults; there was some reduction in firearm-related robberies accompanied by a lesser increase in nonfirearm-related armed robberies; and firearm-related homicides were reduced with no increase in nonfirearm-related homicides. They conclude that the law, in the short term, may have deterred some individuals from carrying or using their firearms, but it did not prevent them from substituting alternative weapons.
Using similar methods, Deutsch and Alt (1977) analyzed police reports of firearm-related assaults, homicide (all types), and armed robbery (including other weapons) for the time period January 1966 through October 1975. The evaluation was designed to detect short-term impacts of the law, as it only included a six-month horizon after the enactment of the law. Deutsch and Alt found a statistically significant 18 percent decrease in gun assaults and a statistically significant 20 percent decrease in armed robberies, but no statistically significant changes in homicide incidents. Hay and McCleary (1979) reanalyzed Deutsch and Alt’s data and suggest that the stochastic components of the time series were not specified correctly and the postintervention time series was too short to permit an accurate specification of the intervention component. Hay and McCleary suggest that the Deutsch and Alt findings are inconclusive. In a rejoinder, Deutsch (1979) critiques the ARIMA model specification choices made by Hay and McCleary in their reanalysis and comments that their research was “wrought with inconsistencies, inaccuracies, and half truths” (p. 327).
In their analysis of the Deutsch and Alt data, Berk and his colleagues (1979) conclude that the law reduced armed robbery, had mixed effects on firearm-related assaults, and had no effects on homicides. In his later analysis, Deutsch (1981) expanded the time series through September 1977 and found that the Bartley-Fox law produced significant reductions in homicide, firearm-related assaults, and armed robbery. The broader methodological lesson learned from this exchange was that identifying appropriate models for evaluation purposes can be a very subjective exercise. As Kleck (1997) suggests, “Experts in [time series] modeling also commonly point out difficulties that even experienced practitioners have in specifying time series models. Specification is very much an art rather than a science, so that different researchers, using the same body of data, can make substantially different, even arbitrary, specification decisions, and, as a result, obtain sharply different results” (p. 354).
Indeed, with such dissimilar findings, it is difficult to specify the effects of the Bartley-Fox law on firearm-related crime. Collectively, this body of research seems to suggest a broad impact on gun crime in Boston. However, it is unclear whether the firearms sentencing enhancement or the mandatory sentence for illegal gun-carrying generated the impact. Kleck (1991) observes that, if one accepts that the Bartley-Fox law worked as a whole, it is risky to infer that other gun-carrying laws would also work, since it may have been a unique constellation of factors in the Boston setting that was responsible for the effect.
Punishment enhancements for firearm-related crimes seem to be justified in sentencing by seriousness considerations, since firearms use in violent crimes increases the likelihood of the victim’s death (Cook and Nagin, 1979). Moreover, there is some evidence to suggest that there should be an incapacitation effect, since gun offenders usually persist in their choice of using a firearm in subsequent crimes (Cook and Nagin, 1979). However, the available research evidence on the deterrent effects of firearms sentencing enhancements on firearm-related crime is mixed, with city-level studies suggesting reductions in firearm-related homicides and possibly other types of firearm-related crime in urban settings (McDowall et al., 1992), as well as nationwide studies suggesting no crime prevention effects at the state level (Marvell and Moody, 1995).
The committee recommends more rigorous study of firearms sentencing enhancement laws at the city level. As Kleck (1997) suggests, state-level analyses suffer from aggregation bias, and lumping heterogeneous jurisdictions into one area could conceal potentially important causal effects at lower levels of aggregation. City-level studies need to engage more rigorous
methods, such as pooled time-series, cross-sectional studies that allow the detection of short-term impacts while controlling for variation in violence levels both across different areas and different times.
PROBLEM-ORIENTED POLICING TO PREVENT FIREARM-RELATED CRIME
Problem-oriented policing may hold some promise for reducing firearm-related violence. Problem-oriented policing works to identify why things are going wrong and to frame responses using a wide variety of approaches. Using an iterative focus on problem identification, analysis, response, evaluation, and adjustment of the response, problem-oriented policing has been applied against a wide variety of crime, fear, and disorder concerns (Goldstein, 1990; Eck and Spelman, 1987; Braga et al., 1999). Our review emphasizes programs specifically aimed at reducing proscribed possession and firearm-related violence.2 Problem-oriented policing programs to reduce firearm-related violence generally focus on reducing the illegal possession, carrying, and use of firearms in gun violence “hot spots” and among violent gun offenders.
While this section categorizes these types of police interventions by whether they are primarily focused on places or offenders, in practice these firearm-related crime prevention strategies overlap. For example, when the police are deployed to prevent gun violence in particular places, they often focus their attention on controlling the illegal gun behaviors of particular individuals in that location. When police efforts are focused on preventing gun violence by likely offenders, such as gang members, they sometimes focus their attention on places, such as gang turf and drug market areas frequented by these individuals. The distinction between a focus on offenders and a focus on places matters less than the idea that the police attempt to reduce crime and violence by strategically focusing on identifiable risks.
Policing Gun Violence Hot Spots
Place-oriented crime prevention strategies have begun to occupy a central role in police crime prevention research and policy (Eck and Weisburd, 1995). This idea developed from the hot-spots crime perspective, which suggests that crime does not occur evenly across urban landscapes; rather, it is concentrated in a relatively few places that generate more than half of all observed criminal events (Pierce et al., 1988; Sherman
et al., 1989; Weisburd et al., 1992). Even in the most crime-ridden neighborhoods, crime appears to cluster at a few discrete locations, and other areas are relatively crime free (Sherman et al., 1989). A number of researchers have argued that many crime problems can be reduced more efficiently if police officers focus their attention on these deviant places (Sherman, 1995; Weisburd, 1997). Sherman and Rogan (1995) suggest three mechanisms through which hot-spots patrol may reduce firearm-related crime in a targeted beat: firearms seized in high firearm-related crime areas may have had significantly higher risk of imminent firearms use in crimes; illegal gun carriers who are arrested may be more frequent gun users; and the visibility of the intensive patrols coupled with increased contacts with citizens may deter gun-carrying by those who are not checked by the police.
Much attention has focused on using place-based policing to reduce gun crime (Sherman, 2001). In this section, we review the evidence from the Kansas City Gun Project and its subsequent replications in Indianapolis and Pittsburgh. All three of these evaluations used place-oriented policing strategies to attempt to confiscate proscribed firearms and prevent crime in gun violence hot spots. We also briefly summarize the anecdotal evidence on the New York Street Crime Unit.
Kansas City Gun Project
The Kansas City Gun Project examined the gun violence prevention effects of proactive patrol and intensive enforcement of firearms laws via safety frisks during traffic stops, plain view searches and seizures, and searches incident to arrests on other charges (Sherman and Rogan, 1995). Over a 6-month period in 1992-1993, the targeted police patrols were conducted in a 10 × 8 block area of Kansas City with a homicide rate 20 times higher than the national average. Simple computer analyses of call and incident data were used to focus police interventions at hot-spot locations. A pair of two-officer cars, working overtime from 7 p.m. to 1 a.m. 7 days a week and not required to answer citizen calls for service, provided extra patrol in the targeted beat.3 Data from the targeted area were compared with data from a beat with nearly identical numbers of drive-by shootings in 1991. The comparison beat received routine levels of police activities.
Comparing the differences in crime rates in the targeted and control communities both before and after the intervention, Sherman and Rogan (1995) assessed the impact of hot-spot policing on firearms seizures and crime. The evaluation concludes that proactive patrols focused on firearms recoveries resulted in a statistically significant 65 percent increase in firearms seizures (29 additional firearms seized) and a statistically significant 49 percent decrease in firearm-related crimes in the target beat area (83 fewer firearm-related crimes); firearms seizures and firearm-related crimes in the comparison beat area did not significantly change (Sherman and Rogan, 1995).4 Furthermore, none of the contiguous beats showed significant increases in firearm-related crime, and two of the contiguous beats reported significant decreases in firearm-related crimes.
Indianapolis Directed Patrol Project
During a 90-day period beginning in July 1997, the Indianapolis Police Department (IPD) implemented a police strategy similar to the Kansas City program (McGarrell et al., 2001). The Indianapolis program tested the effects of two types of directed patrol strategies on firearm-related crime. In the north district, the IPD pursued a directed patrol strategy that sought to prevent firearm-related violence by focusing on suspicious activities and locations. In the east district, the IPD pursued a general deterrence strategy that attempted to prevent firearm-related violence by maximizing the number of vehicle stops in the targeted area. In contrast to the Kansas City study, police activities were not guided by computer analyses of hot-spot locations in either of the targeted areas. Finally, IPD officials worked closely before and during the intervention to secure community support and address concerns (McGarrell et al., 2001). IPD officers were trained to treat citizens with respect and to explain the reasons for the stop.
The evaluation used a pre-post design to determine the effects of the two strategies on firearm-related crime. Both target areas were compared with the same comparison district as well as to citywide crime trends. During the 90-day intervention period, the number of firearms seized in the east district increased by 50 percent, while the north district experienced a modest 8 percent increase (McGarrell et al., 2001). The number of firearms seized in the comparison area decreased by 40 percent. The evaluation revealed that there were statistically significant decreases in firearm-related crime, homicide, aggravated assault with a firearm, and armed robbery in the north district. No statistically significant changes in firearm-related crime were noted in the east district. The evaluation did not reveal any
evidence of immediate spatial displacement of firearm-related crime or significant diffusion of crime control benefits into surrounding areas. It is also noteworthy that not a single citizen complaint was tied to the directed patrol study (McGarrell et al., 2001).
Police Gun Suppression Patrols in Pittsburgh
Over a 14-week period beginning in July 1998, the Pittsburgh Police Department focused on suppressing illegal guns on city streets through the implementation of a special Gun Suppression Patrol program (Cohen and Ludwig, 2003). Two patrol teams of four officers each were assigned to separate police zones experiencing high rates of illegal gun activity. With the aid of crime maps and activity reports on recent shots fired, the patrol teams focused on high-risk times and high-risk places in targeted areas. The patrol teams initiated citizen contacts through traffic stops and “stop and talk” activities with persons on foot. These contacts were used as an opportunity to solicit information and investigate suspicious activities associated with illegal carrying and use of guns. When warranted for officer safety reasons (usually suspicious actions or demeanor), pat-downs for weapons were conducted; when there was reasonable suspicion of criminal activity and an arrest made, these searches sometimes escalated to more thorough checks (Cohen and Ludwig, 2003).
The impact evaluation of the Pittsburgh program used a repeated-differences model. Shots-fired calls for service and firearm-related injuries in the two treatment zones were compared with those in the remaining four police zones in Pittsburgh. The 6-week period between June 7 and July 18, 1998, served as the pre-period, and the 14 weeks between July 19 and October 24 were the post-period. The evaluation found that shots-fired calls for service from residents were reduced by more than 50 percent in one target area, and gunshot injuries were down by nearly 70 percent in the other target area, representing a reduction of 2.5 gunshot injuries weekly in the latter target area (Cohen and Ludwig, 2003).
New York Police Department’s Street Crime Unit
Beginning in 1994, the New York Police Department (NYPD) maintained a special Street Crime Unit that targeted firearm-related violence hot spots and aggressively sought out sources of illegal firearms (Office of Juvenile Justice and Delinquency Prevention, 1999). Between 1994 and 1997, the NYPD made 46,198 gun arrests and confiscated 56,081 firearms. Nonfatal shootings declined by 62 percent between 1993 and 1997 and, in 1998, New York had only 633 homicides, its lowest since 1964 (Office of Juvenile Justice and Delinquency Prevention, 1999). At the same time, the
aggressive policing tactics of the NYPD have been criticized as resulting in increased citizen complaints about police misconduct and abuse of force (Greene, 1999).5 The aggressive gun-oriented policing strategies of the NYPD have not been formally evaluated.6
What Has Been Learned?
The evidence from the three targeted place-based firearm and crime suppression patrols is compelling. All three evaluations are well designed and all reveal the same qualitative conclusion, namely, increased firearms seizures, reductions in crime, and little if any displacement. Moreover, these findings are supported by the larger literature on actual randomized policing experiments, which show place-based policing interventions as having substantial crime control effects (see the National Research Council, 2004).
Despite these encouraging findings, there are several shortcomings in the research information that create uncertainty about the potential efficacy of place-based targeted firearms patrols. At the most basic level, the credibility of the quasi-experimental statistical model rests with whether the underlying comparison group is in fact comparable (Meyer, 1995). In particular, the methodology rests on an assumption that the only important difference between the targeted and control patrol areas is in the intervention. In fact, however, the targeted areas were not chosen at random and were not identical to the comparison patrols. Even if the groups are comparable, these evaluations cannot reveal whether the findings reflect a change from general to targeted policing or a change in resource allocation. In all three evaluations, additional resources were explicitly devoted to the targeted areas. The Kansas City program, for example, included both targeted interventions and additional nighttime patrols. Finally, the interventions were of limited duration and scope, focusing on particular areas at particular points in time. As such, the evaluations may not provide insight into the long-term, large-scale potential of these targeted interventions.
Will hot-spot policing have long-term deterrent effects on gun violence? To what extent will there be geographic substitution of violence? How long will it take criminals to adapt to the new system? Will other forms of crime and violence emerge as police change the focus of their efforts? These are important questions for policy officials who must make decisions about whether and how widely to implement such programs.
Given the early success of these three modest interventions and given the consistency of the basic finding, it would seem worthwhile to learn more about the longer term impacts. Thus, the committee recommends that a sustained and systemic research program be devoted to studying the impact of different place-based gun suppression patrol and targeted policing approaches in general. These evaluations should focus on replicating the existing evidence in different settings, running experimental evaluations, and formalizing and estimating behavioral models of policing and crime. Additional evaluations should assess the longer term impacts, paying particular attention to issues of substitution, adaptation, and deterrence.
Policing Violent Gun Offenders
A small number of chronic offenders generate a disproportionate share of crime. In their seminal study of nearly 10,000 boys in Philadelphia, Wolfgang et al. (1972) revealed that the most active 6 percent of delinquent boys were responsible for more than 50 percent of all delinquent acts committed. The RAND Corporation’s survey of jail and prison inmates in California, Michigan, and Texas revealed that, in all three states, the most frequent 10 percent of active offenders committed some 50 percent of all crimes and 80 percent of crimes were committed by only 20 percent of the criminals (Chaiken and Chaiken, 1982). Moreover, 1 percent of offenders committed crimes at the very high rate of more than 50 serious offenses per year (Rolph et al., 1981).
The observation that a small number of highly active offenders generates a large share of the crime problem is an important insight for law enforcement agencies with limited resources to prevent crime. Many serious urban crime problems, for example gang violence, are driven by groups of these criminally active individuals. Focusing criminal justice attention on a small number of high-risk offenders may be a promising way to control gun violence.
St. Louis Youth Firearm Suppression Program
The Firearm Suppression Program (FSP) sought parental consent to search for and seize the guns of juveniles (Rosenfeld and Decker, 1996). While this program was not explicitly focused on dangerous offenders, it represents a police program to prevent firearm-related violence by disarming a very risky population of potential gun offenders—juveniles. The program was operated by the St. Louis Metropolitan Police Department’s Mobile Reserve Unit, which is a police squad dedicated to responding to pockets of crime and violence throughout St. Louis (Rosenfeld and Decker, 1996). Home searches were initiated on the basis of resident requests for
police service, reports from other police units, and information gained from other investigations. As Rosenfeld and Decker describe, “an innovative feature of the program is its use of a ‘Consent to Search and Seize’ form to secure legal access to the residence. Officers inform the adult resident that the purpose of the program is to confiscate illegal firearms, particularly those belonging to juveniles, without seeking criminal prosecution. The resident is informed that she will not be charged with the illegal possession of a firearm if she signs the consent form” (p. 204). While it was operating, the program generated few complaints from the persons who were subjected to the search, but it received criticism from local representatives of the American Civil Liberties Union, who questioned the possibility of receiving real consent to search when a person is standing face-to-face with two police officers (Rosenfeld and Decker, 1996).
A key component of the program was to respond to problems identified by residents, and the success of the program was reliant on effective police-community relationships. By seeking and acquiring community input into the process of identifying and confiscating guns from juveniles, the St. Louis Metropolitan Police Department developed a model of policing gun violence that put a premium on effective communication and trust with the community not found in most problem-oriented policing projects. As Rosenfeld and Decker (1996) observe, the Firearm Suppression Program was also designed to send a clear message that juvenile firearms possession will not be tolerated by the police or the community because it places individuals at risk and threatens public safety. However, while this program gained national attention for its innovative approach and seemed to be a promising route to disarming juveniles,7 the Mobile Reserve Unit underwent a series of changes that caused the program to be stopped and restarted several times; the subsequent incarnations did not take the same approach as the original program. A rigorous impact evaluation of the original Firearm Suppression Program was not completed.
Boston Gun Project and Operation Ceasefire
The Boston Gun Project was a problem-oriented policing enterprise expressly aimed at taking on a serious, large-scale crime problem—homicide victimization among young people in Boston. Like many large cities in the United States, Boston experienced a large, sudden increase in youth homicide between the late 1980s and early 1990s. The Boston Gun Project proceeded by: (1) assembling an interagency working group of largely line-level criminal justice and other practitioners; (2) applying quantitative and
qualitative research techniques to create an assessment of the nature of and dynamics driving youth violence in Boston; (3) developing an intervention designed to have a substantial, short-term impact on youth homicide; (4) implementing and adapting the intervention; and (5) evaluating the intervention’s impact (Kennedy et al., 1996). The project began in early 1995 and implemented what is now known as the Operation Ceasefire intervention, which began in late spring 1996. While the Boston Gun Project initially focused on firearms and firearm-related violence, the focus evolved as it found that gangs and violent gang offending were central to Boston’s youth gun violence problem. To trigger intervention, any serious violent offending by a gang (knives, blunt instrument beatings) was enough. In practice, however, it was mostly gun offending. Because much of the youth violence epidemic in the 1990s involved firearms and because the Boston Gun Project is cited as a highly effective way to reduce youth firearm-related violence, we devote attention to it in this report.
The project has been extensively described and documented (Kennedy et al., 1996; Kennedy et al., 1997; Kennedy, 1997). Briefly, a working group of law enforcement personnel, youth workers, and researchers diagnosed the youth violence problem in Boston as one of patterned, largely vendetta-like hostilities(“beefs”) among a small population of chronic criminal offenders, and particularly among those involved in some 60 loose, informal, mostly neighborhood-based groups (these groups were called “gangs” in Boston, but were not Chicago- or LA-style gangs, which are much larger and more formally organized). As this diagnosis developed, the focus of the project shifted from its initial framework of juvenile violence and firearm-related violence to gang violence. A central hypothesis of the working group was that a meaningful period of substantially reduced youth violence might serve as a fire-break and result in a relatively long-lasting reduction in future youth violence (Kennedy et al., 1996). The idea was that youth violence in Boston had become a self-sustaining cycle among a relatively small number of youth, with objectively high levels of risk leading to nominally self-protective behavior, such as gun acquisition and use, gang formation, tough street behavior, and the like: behavior that then became an additional input into the cycle of violence (Kennedy et al., 1996). If this cycle could be interrupted, a new equilibrium at a lower level of risk and violence might be established, perhaps without the need for continued high levels of either deterrent or facilitative intervention. The larger hope was that a successful intervention to reduce gang violence in the short term would have a disproportionate, sustainable impact in the long term.
The Operation Ceasefire “pulling-levers” strategy was designed to deter by reaching out directly to gangs, saying explicitly that violence would no longer be tolerated, and backing up that message by “pulling every lever” legally available when violence occurred (Kennedy, 1997). Simultaneously, youth
workers, probation and parole officers, and later churches and other community groups offered gang members services and other kinds of help. The Operation Ceasefire working group delivered this message in formal meetings with gang members, through individual police and probation contacts with gang members, through meetings with inmates of secure juvenile facilities in the city, and through gang outreach workers and activist black clergy. The deterrence message was not a deal with gang members to stop violence. Rather, it was a promise to gang members that violent behavior would evoke an immediate and intense response. If gangs committed other crimes but refrained from violence, the normal workings of police, prosecutors, and the rest of the criminal justice system dealt with these matters. As described below, Operation Ceasefire also attempted to disrupt the illegal supply of firearms to youth by focusing enforcement attention on firearms traffickers.
The evaluation of Operation Ceasefire used a basic one-group time-series design to measure the effects of the intervention on youth homicide and other indicators of nonfatal serious violence in Boston. Braga et al. (2001a, 2001b) found that the Operation Ceasefire intervention was associated with a 63 percent decrease in monthly number of Boston youth homicides, a 32 percent decrease in monthly number of shots-fired calls, a 25 percent decrease in the monthly number of firearm-related assaults, and, in one high-risk police district given special attention in the evaluation, a 44 percent decrease in monthly number of youth firearm-related assault incidents. These reductions associated with Operation Ceasefire persisted when control variables, such as changes in Boston’s employment trends, youth population, and citywide violence trends, were added to the regression models. Furthermore, the basic qualitative results also remained when youth homicide trends in Boston were compared with youth homicide trends in other large U.S. cities. Boston’s significant youth homicide reduction was distinct when compared with youth homicide trends in most major U.S. and New England cities (Braga et al., 2001a, 2001b).8
The dramatic drop in the youth homicide rate in Boston and the associated analysis of Braga et al. (2001a, 2001b) are compelling. Youth homicides in Boston were reduced just after the adoption of Operation Ceasefire.9 However, it is difficult to specify cause and effect. Braga and his
Piehl et al. (1999) examined the youth homicide time series for exogenous structural breaks; these analyses suggest that the maximal break in the series occurred in June 1996—just after the Operation Ceasefire implementation date.
Boston, like many other U.S. cities, experienced a sudden increase in firearm-related violence in 2001. Reported crimes involving firearms increased by over 10 percent between 2000 and 2001 and decreased moderately in 2002 (http://www.ci.boston.ma.us/police/pdfs/dec2003.pdf). McDevitt and his colleagues (2003) suggest that the Boston youth violence problems are dynamic, and the interventions designed to deal with youth violence need to be adjusted appropriately. Since 2001, Boston has been expanding Operation Ceasefire to deal with a wider range of violence problems.
colleagues compare youth homicide before and after the intervention. This type of methodology holds much appeal when an intervention is the only notable event occurring in the time period under study. Observational data from Boston, however, were not derived from an experimental evaluation. To the contrary, during this period of dramatic declines in youth crime throughout the country, there were potentially many levers being pulled in Boston, some controlled by the Operation Ceasefire group and some controlled by outside (and perhaps unobserved) forces. Furthermore, even if all of the determinants of violence except Operation Ceasefire were time invariant, the dynamics that connect enforcement to violence would be complex (these same issues are discussed in National Research Council, 2001). An activity undertaken at a specific place and time presumably does not generate an instant response in violence. And, to the extent that there is a response, it may merely reflect short-term acceleration in the rate of change but not in the steady-state levels in youth crime.
The existing research provides some insight into these potential statistical problems. Braga and his colleagues controlled for demographic shifts, drug market changes, and employment. Moreover, the evaluation shows that the Boston trend is very different from trends in other cities. Kennedy et al. (2001) provide an anecdotal account of the Boston story and Braga et al. (2001a, 2001b) survey the plausibility that other Boston interventions, most notably public health interventions, were associated with the sudden drop. Still, the primary evaluation does allow one to make direct links between key components of the intervention and the subsequent behavior of individuals subjected to the intervention. Many complex factors affect the trajectory of youth violence problems, and, while the there is a strong association between the youth homicide drop and the implementation of Operation Ceasefire, it is very difficult to specify the exact role it played in the reduction of youth homicide in Boston.
In addition to preventing gun violence amongst gangs, Boston’s Operation Ceasefire interagency problem-solving group sought to disrupt the illegal supply of firearms to youth by systematically (Braga et al., 2001a: 199):
Expanding the focus of local, state, and federal authorities to include intrastate trafficking in Massachusetts-sourced guns, in addition to interstate trafficking;
Focusing enforcement attention on traffickers of those makes and calibers of guns most used by gang members, on traffickers of guns showing short time-to-crime, and on traffickers of guns used by the city’s most violent gangs;
Attempting restoration of obliterated serial numbers and subsequent trafficking investigations based on those restorations;
Supporting these enforcement priorities through analysis of crime gun traces generated by the Boston Police Department’s comprehensive tracing of crime guns and by developing leads through systematic debriefing of (especially) arrestees involved with gangs or involved in violent crime.
The Boston supply-side approach was implemented in conjunction with the pulling-levers demand-side strategy to reduce youth violence. The gun trafficking investigations and prosecutions followed the implementation of the pulling-levers strategy, so their effects on firearm-related violence could not be independently established (Braga et al., 2001a). However, the National Institute of Justice, in partnership with the Bureau of Alcohol, Tobacco, and Firearms, recently funded a demonstration program in Los Angeles to examine the effects of disrupting the illegal supply of firearms on the nature of the illegal market and on firearm-related violence (Tita et al., 2003). In addition to addressing the firearm-related violence problem in Los Angeles, this interagency law enforcement project was developed to provide other jurisdictions with guidance on how to analyze and develop appropriate problem-solving interventions to control illegal firearms markets.
Other Applications of the Pulling-Levers Focused Deterrence Approach
After the well-publicized success of Boston’s Operation Ceasefire, a number of jurisdictions began experimenting with these new problem-solving frameworks to prevent gang and group-involved violence. Braga et al. (2002) detail the experiences of Minneapolis (MN), Baltimore (MD), the Boyle Heights section of Los Angeles (CA), Stockton (CA), and Indianapolis (IN) in tailoring the approach to fit their violence problems and operating environments. Although specific tactics sometimes varied across the cities, these programs implemented the basic elements of the original Boston strategy, including the pulling-levers focused deterrence strategy, designed to prevent violence by and among chronic offenders and groups of chronic offenders; the convening of an interagency working group representing a wide range of criminal justice and social service capabilities; and jurisdiction-specific assessments of violence dynamics, perpetrator and victim characteristics, and related issues such as drug market characteristics and patterns of firearms use and acquisition. All were facilitated by a close, more or less real-time partnership between researchers and practitioners. Basic pretest/posttest analyses from these initiatives revealed that these new approaches to the strategic prevention of gang and group-involved violence were associated with reductions in violent crime (Braga
et al., 2002). To date, these replication studies are mostly descriptive in nature.10
What Has Been Learned?
While broad support for the pulling-levers approach may be justified for many reasons, the committee found modest scientific evidence that demonstrates whether these types of targeted policing programs can effectively lower crime and violence. Clearly, there was pronounced and important change in the youth homicide rate in Boston over the period of the intervention, some of which was arguably due to Operation Ceasefire, some due to secular changes in youth homicide, and some due to other (and perhaps unknown) factors. The particular effects of this intervention, however, are unknown. Furthermore, in the committee’s view, the existing data and methods make it difficult to assess how Operation Ceasefire and other similar policing programs affect crime. Researchers cannot hope to credibly control for the many confounders that influence violence and crime using simple time-series comparisons. With similar policing programs being adopted in a number of other areas, there may be opportunities to combine data from these sites to provide more persuasive estimates. Invariably, however, researchers will be confronted with the fact that the programs were not randomly adopted, the trends in violence are influenced by a multitude of factors, and the dynamics of crime and violence are highly complex.
The lack of research on this potentially important intervention is an important shortcoming in the body of knowledge on firearms injury interventions. These programs are widely viewed as effective, but in fact knowledge of how, if at all, they reduce youth crime is limited. Without a much stronger research base, the benefits and harms of these policing interventions remain largely unknown. The committee recommends that a sustained and systematic research program should be conducted to assess the effect of targeted policing aimed at high-risk offenders. Additional insights might be gained by using observational data from different applications, especially if combined with thoughtful behavioral models of policing and crime. An alternative means of assessing the impact of these types of targeted policing interventions would be to run randomized experiments, similar in spirit to those described above. Using this framework, one might hope to disentangle the effects of the various levers and more generally assess the effectiveness of these targeted policing programs.