Title: National Evaluation of the Youth Firearms Violence Initiative.
Series: Research in Brief
Author: Terence Dunworth
Published: National Institute of Justice, November 2000
Subject: Community policing
22 pages
54,000 bytes

-------------------------------

Figures, charts, forms, and tables are not included in this ASCII plain-text
file. To view this document in its entirety, download the Adobe Acrobat
graphic file available from this Web site or order a print copy from NCJRS
at 800-851-3420 (877-712-9279 for TTY users).

-------------------------------

U.S. Department of Justice
Office of Justice Programs
National Institute of Justice

National Institute of Justice
Research in Brief
Julie E. Samuels, Acting Director
November 2000

National Evaluation of the Youth Firearms Violence Initiative

By Terence Dunworth

-------------------------------

Support for this research was provided through a transfer of funds to NIJ
from the Office of Community Oriented Policing Services.

-------------------------------

Issues and Findings

Discussed in this Brief: The national evaluation of the Youth Firearms
Violence Initiative (YFVI), a program initiated by the Office of
Community Oriented Policing Services (COPS) to combat the rise of
juvenile firearms violence. COPS provided up to $1 million to 10
participating cities to fund interventions that employed community
policing approaches to decrease the number of violent firearms crimes
committed by youths, including gang- and drug-related offenses.

Key issues: From 1985 to 1994, the rate of violent criminal acts
committed by juveniles rose sharply at a time when violence committed by
adults was declining. During that time period, while the adult homicide
rate fell by 25 percent, the homicide rate for 18- to 24-year-olds doubled,
and the rate for 15- to 17-year-olds more than tripled. More specifically,
juvenile homicides committed with a handgun more than doubled, while
juvenile homicides committed with other weapons remained essentially
unchanged.

Key findings: Among the evaluation's findings were the following:

--A dedicated unit may exert a greater effect on gun-related crime than a
unit that applies traditional tactics and uses patrol officers on a rotating
basis.

--When employed as part of YFVI, traditional enforcement tactics did not
produce significant changes in firearms violence levels. 

--Cooperation with other law enforcement agencies and community
organizations and representatives was a key factor in effective
implementation of firearms violence control and prevention strategies.

--Proactive arrest policies, focused on gun-related offenses, were shown to
have a consistent measurable association with subsequent gun-related
crime.

--Most of the participating departments returned to traditional policing
approaches when Federal funding ended. In this sense, YFVI did not
change the way departments conducted their business, and YFVI strategies
were not, in general, institutionalized.

--Most departments found it difficult to implement the geographic
information systems called for by the initiative. However, all five
intensive sites maintained computerized files on incidents (arrests and
crimes), permitting geographic analysis after incident addresses were
parsed. 

--Not all departments implemented programs that focused on the
objectives of the YFVI. This suggests that Federal agencies will
sometimes need to play a strong role in ensuring that grantees address and
meet the strategic intent of the initiatives they fund. The challenge is for
the Federal agency to do this while simultaneously promoting local
generation and definition of programmatic activities.

Target audience: State and local law enforcement, probation, and parole
officials; policymakers and planning officials; court administrators; and
researchers.

-------------------------------

Between 1985 and 1994, violent criminal acts committed by juveniles
grew at an alarming pace in the United States. This trend diverged
significantly from adult patterns. From 1985 to 1994, for instance, the
homicide rate for adults age 25 and older declined 25 percent. During the
same period, the homicide rate for 18- to 24-year-olds doubled, and the
rate of homicides committed by teenagers ages 15 to 17 tripled. (See
exhibit 1.)

Handguns became the weapon of choice for juveniles. Exhibit 2 shows
that during this time period, the number of homicides committed by
juveniles with handguns more than doubled, while the number of juvenile
homicides involving other weapons remained essentially constant.

The Youth Firearms Violence Initiative (YFVI) was launched in 1995 by
the U.S. Department of Justice's Office of Community Oriented Policing
Services (COPS). It provided up to $1 million to the police departments of
10 participating cities (Baltimore, Maryland; Birmingham, Alabama;
Bridgeport, Connecticut; Cleveland, Ohio; Inglewood, California;
Milwaukee, Wisconsin; Richmond, Virginia; Salinas, California; San
Antonio, Texas; and Seattle, Washington) to fund interventions directed at
combating the rise of youth firearms violence. The initiative encouraged
these jurisdictions to employ community policing approaches to develop
or enhance youth-focused programs designed to decrease the number of
violent firearms crimes, reduce the number of firearms-related gang
offenses, and reduce the number of firearms-related drug offenses. 

YFVI was conceived at a time when juvenile violence was ascending and
seemed little influenced by any previously attempted interventions.
Although some strategies seemed promising (the Boston Gun Project[1]
and the St. Louis "Knock and Talk" program,[2] for example), they were
not in general use, and the criminal justice community expressed little
confidence at that time that the solutions to the problem had been
identified. YFVI was an attempt to rectify that deficiency.

COPS provided general programmatic guidelines to the 10 departments
but requested action plans that reflected local needs. The cities proposed a
mix of enforcement programs (both street level and school based),
prevention programs, and information systems enhancements. Exhibit 3
shows budget allocations by category for each site, as set forth in the final
plans. Overall, about 60 percent of the grant was budgeted for law
enforcement staff, and most of this was devoted to police personnel,
through either overtime expenses or the creation of new positions. A little
less than one-fourth of YFVI funds was budgeted for local evaluation,
civilian consultants (e.g., for training), and community-based
organizations and activities. Nearly 15 percent--more than $1 million
collectively--was budgeted to purchase computer hardware and software
or to develop information systems.

The national evaluation

The National Institute of Justice (NIJ) sponsored the national evaluation of
YFVI (from the fall of 1995 through the summer of 1998) to assess the
impact of the sites' strategies for addressing youth firearms violence, to
describe how the sites implemented these strategies, and to inform other
jurisdictions about the successes and failures of these strategies. The
evaluation solicitation specified that there should be five "intensive" and
five "nonintensive" sites. In the intensive sites, impact and intensive
process evaluations would be conducted; in the nonintensive sites, only
summary process evaluations would be conducted.[3] The impact
evaluations would determine the impact of YFVI on levels of youth
firearm violence, while the process evaluations would focus on program
implementation and organization.

Specific issues that were investigated in the process evaluations at all 10
sites included:

--Program development: How and why the sites selected their strategies,
their target selection process, implementation problems encountered
during the intervention, and changes made to the program during the
intervention.

--Officer training: The content and focus of YFVI-related training
implemented at the sites.

--Team selection: The process for selecting supervisors and officers for
YFVI interventions and any problems caused by the selection process.

--Computer systems support: The computer systems used to support YFVI
operations and the overall utility of these systems for such an initiative.

--Strategies and tactics: The types of enforcement, community policing,
prevention, and other activities undertaken with YFVI grant funds.

--Program results: What the YFVI-funded police officers and other
program elements accomplished, such as the number of arrests made and
the number of guns seized.

--Local evaluation role: The types of evaluation activities undertaken and
an overall assessment of the appropriateness of the evaluation approaches.

This information was obtained through a review of program materials
(project-related documents, such as the site's proposal to COPS, quarterly
progress reports, interim and final reports, and other internal documents);
periodic telephone and onsite interviews with police supervisors and
officers participating in the YFVI initiative; ridealongs with YFVI
officers; interviews with local evaluators; and an analysis of summary data
provided by the police department (YFVI enforcement schedules and
activity logs, for example).

COPS, NIJ, and Abt Associates Inc. determined which sites would be
designated as intensive and nonintensive, based on the following criteria:

--Evaluability, or the likelihood a department's strategies would have a
measurable impact within the evaluation period: The evaluability potential
was considered lower for departments that focused on prevention and
education (because the window of observation was too short) or that had
diffused or short-term target areas (too scattered an effect in the former,
not enough time in the latter).

--Information systems status: Change in the level of gun-related crime was
a critical component of the impact evaluation. Thus, complete
computerized records spanning several years, with accurate ages of
offenders and suspects and specific offense codes for firearms-related
crimes, increased the likelihood of successful intensive evaluation.

--YFVI implementation status: It was clear that some sites might
experience significant delays in starting their YFVI program. Given the
short evaluation time period for each site, it was critical that the five
intensive sites have programs up and running as soon as possible.

--Nature of intervention: Some sites implemented primarily police-based
enforcement strategies, while other sites emphasized prevention,
community policing, and other intervention approaches. Since the
evaluation was to focus on a broad range of strategies, the intensive sites
should include both enforcement and prevention tactics, despite the greater
difficulty of evaluating the latter.

Based on these criteria, Baltimore, Cleveland, Inglewood, Salinas, and San
Antonio were selected as the intensive sites. Birmingham, Bridgeport,
Milwaukee, Richmond, and Seattle became the nonintensive sites. This
Research in Brief examines the experience of the intensive sites.[4]

Strategies and tactics

The strategies and tactics proposed by the YFVI departments focused on
three areas: streets, schools, and communities. Each department employed
geographic information systems and crime analysis to support all three
areas. Exhibit 4 summarizes the main program elements set up by each of
the five intensive sites.

COPS encouraged the police departments to:

--Work in conjunction with other city agencies to promote education,
prevention, and intervention programs related to handguns and handgun
safety.

--Develop community-based programs focused on youth handgun
violence.

--Develop programs involving and assisting families in addressing youth
handgun problems.

Furthermore, COPS encouraged jurisdictions to employ community
policing strategies to:

--Implement programs specifically designed to decrease the number of
violent firearms crimes committed by youths.

--Develop or enhance programs aimed at reducing firearms-related gang
and drug offenses.

--Address the root causes of youth firearms violence in specific geographic
areas.

Traditional enforcement operations. As noted earlier, most sites used
street-based operations that featured traditional enforcement tactics that
the departments perceived to be in compliance with the programmatic
preferences that COPS had communicated at the outset of the initiative.
Four of the five intensive evaluation sites--Baltimore, Inglewood, Salinas,
and San Antonio--either set up new enforcement units or expanded
existing teams. All used traditional surveillance and intelligence-gathering
techniques to identify targets and to focus the teams' efforts. Actual
operations, for the most part, involved location-specific, street-level
activities. With the exception of Salinas, which adopted a citywide
approach, YFVI teams targeted specific neighborhoods within their cities.
During the life of the initiative, the teams adjusted their targets based on
their perceived effects on the problem or the likelihood that firearms
would be present at a particular location at a particular point in time. To
make such judgments, the teams used a combination of intelligence and
analyses of their own past performance.

Perhaps the greatest variation among departments was in the way the
special teams were organized and staffed. At one extreme was the fully
dedicated unit where participating officers were relieved of all regular
patrol duties except in emergencies (as in Salinas); the other was the
rotating assignment of officers to the team on a temporary overtime basis
(as in San Antonio).[5]

In Salinas, roughly 10 percent of the sworn complement of 160 officers
was assigned to a Violence Suppression Unit (VSU). They were led by a
lieutenant, and two sergeants each managed a team of eight officers. The
two teams worked a 4-day week, with an overlap day used for intelligence
sharing, cross-team debriefing, and planning. The VSU was housed in a
location separate from (though close to) the main police station, and it
operated independently. Team members were relieved of all normal patrol
functions, including response to calls for service, and were given offsite
training in other cities. Team leaders planned activities in conjunction with
team members and developed relatively long-range plans to address the
youth firearms problem. The team established, monitored, and maintained
an intelligence system relating specifically to youths.

This approach led to a high level of acceptance by the team members and
produced a highly focused effort. Members rotated out of the unit only for
cause, became very knowledgeable about the problems they addressed,
and came to know and be known by youths in the city. This created a
degree of continuity that was highly valued by team members.

The dedicated approach establishes an elite unit within a department,
risking a negative reaction from nonmembers. This problem was greatly
minimized in Salinas because regular patrol officers were offered overtime
to backfill the patrol functions team members no longer performed.
Furthermore, VSU leadership took pains to keep the department informed
about team activities, and cooperative activities involving the team and
patrol units were promoted whenever possible.

The San Antonio Weapons Recovery and Tracking Team (WRAT)
functioned at the opposite pole. Union rules mandated the equitable
rotation of overtime across officers, producing a situation in which a
different set of officers might work on WRAT from day to day. This
model can be considered positive for officer equity and opportunity, but it
limited the continuity of the team's activities. Five officers had limited and
sporadic exposure to the initiative's undertakings, and planning and
organization proceeded without much input from them. YFVI activities in
San Antonio, therefore, tended to be more reactive than in Salinas. This
made WRAT more like traditional enforcement and less like community-
or problem-oriented policing.

Beyond traditional enforcement. Some departments supplemented
traditional enforcement approaches with other techniques, some of which
were enforcement oriented but which differed from the street-level
approach discussed above. Others were community or school based and
embodied a more preventive strategy.

The Inglewood Police Department partnered with the probation
department, which detailed a full-time officer to work from police offices
to scrutinize gang members on probation. This partnership was extremely
successful, and a significant number of violations were detected.
Inglewood also established the Strategy Against Gang Environments
(SAGE) program, which filed contempt-of-court sanctions against gang
members who violated civil injunctions against assembly with other gang
members. This program involved close cooperation between the police
department and the district attorney's office, with the latter drafting and
obtaining the civil injunction from the courts. It was difficult and time
consuming to set up, it garnered little community support, and the
eventual effects were uncertain. Inglewood police officers also worked
with community leaders to set up the Rites-of-Passage Mentoring Program
to teach youths civic values, self-esteem, and conflict mediation.

In Cleveland, the initiative funded a Residential Area Policing Program
(RAPP). In a small number of troubled, violent neighborhoods, residences
once used as crack houses or for other illicit purposes were converted into
police substations. Community residents met there to coordinate such
activities as neighborhood cleanups and to communicate community
problems to the police, and local youths would drop in to talk with
officers. The houses were staffed around the clock by Cleveland officers
for a period of 90 days. RAPP Houses supported by YFVI funds were set
up consecutively in three different neighborhoods. Community residents
responded favorably, and feedback was positive. When the police
department proposed shutting down the RAPP Houses, communities
pressed successfully for their continued operation.

Baltimore also initiated school-based activities, deploying a small number
of officers in schools on a more or less dedicated basis. The officers
emphasized educating students about risk, supported a specific criminal
justice curriculum, and served as counselors and advisers as needed.
Baltimore extended these activities to the community by establishing
community resource centers and providing information and support to
students and families relating to curfew violations and truancy.

Outputs and impacts

The YFVI assessment had a number of dimensions, including examining
the effectiveness of the organization and structure of the interventions--
that is, a qualitative review of the processes and procedures that were
employed. Other dimensions involved measuring results generated by the
interventions and their impact on the problem of youth violence. To
evaluate results, data on the number of arrests made and guns seized by
YFVI teams were considered output measures. To evaluate the possible
impact on the problem, crime trends were analyzed in the four intensive
sites that emphasized enforcement interventions; in Salinas, a model of the
relationship between arrest activity and subsequent gun crime levels was
developed. 

Output: Arrests made and guns seized. Exhibit 5 presents the reported
number of arrests made and guns seized in four intensive sites. No
numbers are reported for Cleveland because the RAPP Houses did not
make arrests and seizures a goal, and the RAPP neighborhoods were too
small to yield useful statistical information.

The numbers vary from site to site despite relatively similar levels of
Federal funding. However, although such counts need to be reported,
caution must be exercised before these output measures are used to draw
conclusions about program effectiveness in any of the sites. There are
several reasons for such caution.

First, the correspondence of arrests and seizures to YFVI activities varies
by site. Some sites implemented effective tracking systems to capture
YFVI enforcement information, but others did not. Salinas data, for
example, accurately represent the VSU activities because the counts were
sufficiently detailed that individual officers' activities could be identified
from computerized department records. In other sites, however, the
information presented in exhibit 5 was not so easy to interpret. The
Inglewood numbers, for instance, are probably an undercount because they
reflect activity only in the primary target area (Darby-Dixon) and do not
capture YFVI team operations outside that area. San Antonio numbers
were probably an overstatement because they reported arrests and seizures
made by officers outside, as well as inside, the YFVI unit.

Second, the scope of the data available for review did not permit an
assessment of the relative quality of the arrests made. This information
would be helpful. For example, an arrest of a youthful repeat felon may
have a significantly greater effect on public safety than a first-time arrest
of an adult offender for a minor infraction. 

Third, large numbers of arrests and seizures cannot necessarily be
considered a sign of success, and few arrests and seizures do not
necessarily indicate failure. Geographically focused enforcement efforts
viewed as successful typically start out with high arrest figures; then, the
number of arrests decreases significantly. An intervention that emphasizes
deterrence through frequent contact and involvement with potential
youthful offenders may produce fewer arrests than a traditional
enforcement approach. It might be characterized by a greater number of
field interviews, for instance. Yet it might be more effective in suppressing
violence than enforcement efforts that emphasized arrests and seizures.

Fourth, some sites, such as San Antonio, used tactics specifically designed
and implemented to make arrests and seize guns, while other sites
deliberately implemented tactics designed for other purposes. In
Cleveland, for example, the RAPP House officers focused on community
interaction and stabilizing the neighborhood instead of on arrests and
seizures.

Finally, a citywide enforcement strategy--such as the approach
implemented in San Antonio--was more likely to yield larger numbers of
arrests and seizures than a strategy that focused on a small geographic
area, where the extended presence of YFVI officers would undoubtedly
reduce the likelihood that persons would carry guns on their person.
Whether YFVI officers responded to citizen calls for service or if they
were freed from that responsibility and could focus exclusively on their
proactive enforcement efforts was another related factor.

Two sites--Salinas and San Antonio--experimented with different tactics
to determine which yielded the greatest numbers of arrests and seizures.
San Antonio, for example, employed five major tactics: search warrants,
saturation patrol, directed patrol, knock and talk, and bar checks. Salinas
used eight major tactics: profile enforcement, suppression enforcement,
probation searches, surveillance enforcement, search warrants, arrest
warrants, informant development, and gang certification. A productivity
measure for each tactic was produced based on the amount of time devoted
to the tactics and the number of guns seized using each tactic. Executing
search warrants was by far the most productive tactic in both sites. In
Salinas, this tactic was 4.5 times more likely to result in a gun seizure than
any other tactic. In San Antonio, the cost (in terms of officer hours) per
firearm seized via search warrants was about 10 times less than the cost
per seizure of any other tactic. Additional details on these figures are
available in the Salinas and San Antonio case studies (see "Supplementary
YFVI Reports"). 

Impact: Gun crime trends. To assess gun crime trends, incident-level
police data were obtained from the five intensive sites from January 1993
until the middle of 1997. These data provided access to facts about crimes
and arrests, including the time and place of occurrence, the crime
associated with the incident, and the suspects (if any). This information
made it possible to map the incidents (identifying which ones took place in
target areas), to derive or estimate the age of the perpetrators (designating
incidents involving juveniles), and to establish a baseline for at least 2
years prior to the commencement of YFVI (facilitating comparisons over
time). Standardized procedures for classifying the data were developed
and applied to all sites. (See "Crime Classification.")

In each of the five cities, reported gun crimes declined after YFVI was
implemented. Reductions also took place in all but one of the target areas
(see exhibit 6). Citywide decreases were greatest in San Antonio (41
percent), followed by Inglewood (23 percent), Cleveland (15 percent),
Salinas (11 percent), and Baltimore (2 percent).

Target-area reductions were greatest in Darby-Dixon in Inglewood (49
percent), Cleveland RAPP House areas (38 percent), and San Antonio (37
percent). In Baltimore, the Cherry Hill target area experienced no change,
while Park Heights declined by 8 percent. In Salinas, where the initiative
was conducted citywide, the target-area concept is irrelevant. 

When these broad comparisons are made, the target-area decreases in gun
crime were considerably greater than citywide decreases in three of the
target areas (Cleveland RAPP Houses, Darby-Dixon in Inglewood, and
Park Heights in Baltimore). Though San Antonio's target areas showed a
significant decline (37 percent), the amount was less than that which
occurred in the city at large, and Cherry Hill in Baltimore showed no
change at all.[6]

These comparisons are consistent with the view that the interventions had
an effect in some of the sites, even though they do not demonstrate
causality. Comparing the percentages of gun crimes involving youths
before and after YFVI provides further reason to be cautious. Though
encouraging declines in the percentage of youth involvement in violent
acts occurred in three of the five locations--from 70 percent to 64 percent
in Darby-Dixon in Inglewood, from 66 percent to 52 percent in the four
San Antonio target areas, and from 69 percent to 63 percent around the
Cleveland RAPP Houses--these reductions are statistically significant only
in San Antonio (due to the small number of recorded offenses in the other
locations). The reduction in Baltimore's Park Heights section (from 54
percent to 52 percent) was significant, but small, and was no greater than
the citywide reduction. And finally, both Cherry Hill (Baltimore) and
Salinas experienced virtually no decline at all in this measure.

Charting the trends in general gun crime across the 5 years for which there
are data conveys additional information. Exhibit 7 displays the trends for
four of the sites.[7]

Prior to YFVI implementation, gun crime patterns in the target areas were
similar to those in the surrounding city or county. Levels of gun crime rose
in the city of Salinas and in Monterey County until late summer 1996,
after which time they declined. Gun crime increased in San Antonio city
and the target areas until the beginning of 1996 and then fell sharply. Gun
crime in the other two sites gradually decreased in both the target areas
and the cities at large. 

The similarity between the large area and small area patterns suggests that
strong societal forces exerted a general influence on gun crime trends and
that these worked similarly in most places. Such forces were difficult to
identify conceptually, let alone measure and monitor definitively.
Consequently, although it was easy to make this imputation, it was much
more difficult to document it.[8]

Though the observed trends do not demonstrate a causal connection
between the initiative and the declines, they offer some supporting
evidence for the notion that the initiative helped produce the desired
effects. If, for instance, target areas showed no greater decline than
surrounding areas or increased at a time when the general environment
declined, then this would constitute supporting evidence for the
interpretation that the initiative had little or no effect. The most dramatic
change occurred in Inglewood, where gun crimes in the Darby-Dixon
target area dipped sharply during the YFVI period and then rebounded to
1994 levels after YFVI ended. Salinas also experienced a proportionally
greater reduction than surrounding Monterey County; again, this took
place during the YFVI period. In both cities, this evidence supports the
notion of a YFVI impact on gun crime.

In San Antonio and Baltimore, target-area changes were either reductions
similar in scope to the city at large or increases at a time when the city was
declining. In these two cities, the gun crime trends did not support the idea
that YFVI had an effect on gun crime.

These interpretations rest on a number of assumptions about the YFVI
intervention: (1) that the underlying target-area trends would have
continued to be like the surrounding area trends during and after the YFVI
timeframe; (2) that simple counts of gun crimes are a reasonable measure
of the impact of interventions of this kind; and (3) that the primary effects
of the intervention occurred during the measurement period, rather than
later. It is easy to see how these assumptions might be invalid. For
example, the trends almost certainly do not capture possible qualitative
YFVI effects, such as a reduced potential for younger children to become
involved in gun crimes after the evaluation period ended. Also, it is
important to remember that the number of gun crimes committed per week
in all of these cities was quite small; consequently, a small change from
week to week produced a large effect on the chart.

Effects of arrests on gun crimes in Salinas. To explore the effect of
gun-related arrests on the ensuing level of gun-related crimes, the research
team analyzed 210 weeks of Salinas gun crimes and gun-related arrests.
Using data from January 1993 through December 1996, researchers
estimated the effect of intensified police activity directed toward
gun-related crimes committed by youths. The number of gun-related
arrests (in part attributable to YFVI) was considered a surrogate for the
level of police activity. The variables used were weekly counts of
gun-related arrests and gun-related crimes, as well as a violent crime index
for surrounding Monterey County, excluding Salinas (countywide gun
crime data were unavailable, so violent crime was substituted).[9]

The results indicated that gun-related crimes in Salinas were positively
related to Monterey County's crime index (p = .028) and negatively related
to previous gun-related arrests in Salinas (p = .051). Specifically, a
10-percentage-point increase in Monterey County's crime index was
associated with one additional gun crime in Salinas, and an increase of
five arrests in Salinas was followed by one fewer gun crime in that town.

The pace at which the effect of past arrests unfolded is depicted in exhibit
8.

The graph indicates that 40 percent of the effects of increased arrest
activity occurred within 1 week after an arrest, 64 percent (the sum of 40
percent and 24 percent) occurred within 2 weeks, and so on. Because more
than 95 percent of the eventual effect occurred within 6 weeks, then, for
practical purposes, this can be considered the time interval within which
the eventual total effect of arrests on gun crime will be made.

This model suggests with a high degree of statistical confidence that, in
Salinas at least, enforcement directed proactively at firearms possession
and use has a measurable effect on subsequent firearms crime. This
implies that such activities likely had a quick and salutary impact on
crime. Furthermore, this relationship persisted over a fairly long period of
time (4 years in this city). Nevertheless, the potential magnitude of the
impact is obviously finite and is bound to diminish at some point. That is,
the marginal impact of additional units of effort (say, one 8-hour shift for
one officer) will decline as more shifts are added. In addition, other cities
may not have the same experience as Salinas. In fact, during the YFVI
timeframe, most did not. This may be due to the somewhat unique nature
of the Salinas situation--a relatively small, geographically distinct city
surrounded by agricultural land, in which 10 percent of the police
department's entire sworn complement was dedicated to YFVI to the
exclusion of all other policing responsibilities (except emergencies). To
match this, a city such as San Antonio, for example, would have needed to
devote about 150 officers to YFVI on a full-time basis. Such a level of
activity would have been impossible to support through YFVI because
each city received roughly the same amount of Federal support, regardless
of its size.

Despite these caveats, the Salinas experience provides strong empirical
affirmation of what many chiefs and officers believe intuitively: A quick
"knock down" effect can be achieved by intensified enforcement. The
diminishing marginal utility of intensified enforcement should be
investigated further. Given the inevitable constraints on police department
revenues, it would be useful, for instance, to calculate the optimal
periodicity and targeting of intensified enforcement--that is, where, when,
and how often it should be undertaken. Also, it is worth investigating
whether there is displacement to other geographic areas or a resurgence to
earlier levels in the same area.

Residual effects of the initiative

Programs such as YFVI raise a critical question: What happens when the
Federal funding ends? Of particular interest are the possible effects YFVI
programs have on the police departments themselves. That is, what is the
prognosis for the YFVI program to have a long-term institutional impact
on the police hosting the intervention?

Modification in orientation. Police officers and supervisors in most YFVI
sites commented on their department's use of overtime to deal with crime
more effectively. Relieving YFVI officers of normal calls-for-service
response requirements--getting them "off the radio"--was an important
objective in some cities. The general idea of many of these programs was
to pay overtime to the YFVI officers (or the replacements who backfilled
their positions) while they were in the YFVI program. The police agency
benefited because it increased police presence in the targeted areas, and
this presence was not created at the expense of 911 or other call-taking
systems. In this instance, overtime provided organizational flexibility,
which was devoted to youth and firearms problems.

Many police officials said unions and other officers were much more
accepting of the program when overtime money was made available to a
wider number of police officers through the backfill process. To some
extent, the potentially negative effects of the YFVI officers' "elite" status
were overcome by paying other officers to work in the positions vacated
by the YFVI team.

The logic offered here is straightforward: Federal funding can be a force
multiplier. More officer hours can be spent on the street without increasing
the number of officers in the department. However, there is an important
complication. Implicitly, the Federal Government and the departments
themselves say such efforts as YFVI would not be feasible--even if
desirable and valuable--if additional resources were not provided through
Federal support. Departmental capacity is locally perceived as fully
engaged in the operations being conducted prior to the provision of
Federal funds. 

This presumption seems to inhibit the likelihood of program
institutionalization and may undermine several problem-oriented and
community-oriented policing premises. For example, such a presumption
suggests that community policing and problem solving are luxuries,
affordable only after the basic functions of policing are accomplished.
This implies that a combination of budgetary constraints and political
pressures to deal with calls for service impedes a department's ability to
modify its major response systems. Rather, a department would continue
using city budgets to finance regular patrols and employ supplementary
State and Federal dollars for special programs. Absent the latter, such
initiatives as YFVI might, therefore, be impossible to launch. Such an
interpretation seems to provide additional support, both normative and
economic, for continuing traditional patrol responses.

This arrangement also implies that the cost of providing problem-solving
services becomes roughly 1.5 times that normally incurred by the police
(because the services are paid through overtime). In this sense, the cost of
police services in a jurisdiction receiving Federal funding will go up, in
both a relative and an absolute sense, even though the increases are
"shared" by the Federal taxpayer. If the YFVI program is established to
increase problem-solving applications by police departments, the model of
creating overtime to do so may be counterproductive to the program's
long-range goals.

Consequently, the effects of such programs as YFVI may be restricted to a
short-term subsidy of traditional police patrol responses, unless a
department either adjusts its strategic and tactical responses (learns from
the initiative) or identifies other funding sources to fill the gap when
Federal funding ends. This view has negative implications for continuing
and institutionalizing a program--a goal of YFVI and of most Federal
support programs (the Byrne Formula Grant Program and the Local Law
Enforcement Block Grant Program, for example). Without sustained
resources from a source outside the regular city budget, it seems unlikely
that such institutionalization will occur in many places, at least not on the
scale of the funded program.

Noteworthy approaches. For the most part, the police tactics employed in
YFVI were largely adapted from the patrol model. These included heavily
patrolling selected areas of the city, making traffic stops, and conducting
field interrogations. In this sense, despite the reservations mentioned
above, most sites implemented a significant enforcement effort aimed at
youth firearms violence. In addition, at least three sites (Cleveland,
Inglewood, and Salinas) undertook strategic innovation as a direct
consequence of YFVI. 

The Cleveland RAPP Houses created a new approach to dealing with
distressed communities, an approach that involved more than just a
neighborhood substation. The RAPP Houses differ from substations
because a deliberate outreach effort was undertaken to draw area youths
into direct contact with police officers in a relaxed setting. It is extremely
difficult to evaluate the effect of this kind of program because the
behavioral change it might induce is specific to individuals and long term.
Followup with particular individuals might be necessary to assess how the
RAPP House approach affected them. However, this was beyond the scope
of the evaluation and the timeframe of the initiative. The researchers
ascertained that community reaction was so positive that residents lobbied
for continuing the first RAPP House when it was scheduled to close. In
addition, participating officers were very supportive and uniformly
appeared to favor the strategy. It was also possible to affirm that the
program would continue at some level when the initiative ended,
suggesting the approach had been institutionalized within the command
structure of the police department and city government.

The Inglewood initiative was notable for its combination of enforcement,
collaboration with other criminal justice agencies (probation and the
district attorney), and creation of community programs (Rites-of-Passage).
The police-probation collaboration was particularly effective. Police and
probation officials alike credited the funding of a deputy probation officer
and housing that officer in the police department as contributing
significantly to YFVI's effectiveness. Whether this arrangement could be
continued was not clear at the time the evaluation ended, but doing so
would be highly productive.

The Salinas intervention was unique because it created a dedicated team of
officers working full time on YFVI, with no responsibilities for normal
patrol or routine call response. (Emergency responses were still
undertaken.) It was also unique for the magnitude of the effort relative to
the total size of the force. Salinas has roughly 160 sworn positions. The
VSU had 1 lieutenant, 2 full-time sergeants, and 16 full-time officers--
more than 10 percent of the department's entire force. Finally, its focus
was strategically appropriate for a YFVI response: It targeted youths in a
continuous and deliberate manner.

Summary

The evidence from Inglewood and Salinas supports the view that the YFVI
intervention coincided with a significant drop in the level of gun crime in
both places. In Inglewood, not only did gun crimes drop dramatically
during the YFVI implementation period, but after the initiative concluded,
gun crimes returned to their preinitiative levels. Reductions also took place
in Salinas and continued through the end of the observation period. One
difference between the two sites is that although Federal funding ended in
Salinas at about the same time as in Inglewood, the Salinas Police
Department continued the VSU program at essentially the same level.
Inglewood, on the other hand, terminated its police-probation
collaboration due to lack of funds. The implication is that intervention
most likely reduced gun crime in both places. The comparison of
target-area experience with citywide trends supports this interpretation.

In the other intensive sites, YFVI interventions did not appear to produce
these results. The reasons for this finding are not completely clear,
although the importance of Salinas' dedicated and focused Violence
Suppression Unit and Inglewood's highly successful partnership between
the police department and the probation office should not be
underestimated.[10]

In conclusion, the results of the YFVI effort seemed to depend on how
departments conceived the initiative and the consistency of their focus
throughout its life. Given the problems experienced by some departments
with conceptualization and implementation, variability in outcomes was to
be expected. These observations suggest the Federal Government could
play a stronger role in ensuring that participating departments adhere to
the strategic goals and objectives in future initiatives of this kind. Also,
funding agencies should adopt a timetable that more closely matches the
exigencies of the world in which most departments operate. In particular,
it is desirable for the Federal agency to ensure that local recipients do not
view an initiative as simply a transfer of Federal funds to another level of
government. Simultaneously, care must be taken not to impose Federal
solutions on local problems; that is, the strategies and tactics that are
employed must be locally developed.

Notes

1. The Boston Gun Project was a problem-solving effort to reduce gang
activity and interrupt the self-sustaining cycle of fear and weapons
acquisition and use. The methods included (1) a use-reduction strategy that
employed both traditional and new technological gun tracing to identify
and interrupt gun flow to youths and (2) a deterrence strategy that
communicated to youths the severe criminal consequences they would
face if they were caught with firearms in their possession. 

2. The St. Louis "Knock and Talk" program involved a collaboration
between the police and community to target firearms possession by
youths. Police and probation officers made informal visits to the homes of
youths suspected of possessing guns. The problem was discussed with
parents and may have included a request to permit a search for guns.
Respondents were immunized from prosecution based on the information
they gave, though not from prosecution based on uncovered evidence (if a
gun used in a crime was found, for example).

3. The process evaluations are reported in detail in the 10 case studies and
the cross-site report. See "Supplementary YFVI Reports."

4. The 10 sites were split into 2 equal groups for budgetary reasons. No
comparative ranking of site interventions or results was implied by the
actual selection. Case studies for all 10 sites were produced by the
evaluation and can be located through the National Criminal Justice
Research Service. See "Supplementary YFVI Reports."

5. Additional details on these figures are available in the Salinas and San
Antonio case studies. See "Supplementary YFVI Reports."

6. The declines that occurred were all statistically significant beyond the
.05 level with the exception of the Cleveland RAPP Houses, where the
low number of gun crimes leads to a p = .14. 

7. Cleveland is excluded because the level of gun crime in the RAPP
House areas was too small for charts to be useful. In the other sites,
target-area data are compared to citywide data, except in the case of
Salinas, where the county was substituted because the intervention was
citywide. Also, violent crime in general was substituted for gun crime in
Salinas because data on the latter were not available. 

8. The trends documented in the charts are, in fact, similar to national
trends. These show that significant expansions of gun crime occurred
between the mid-1980s and the mid-1990s, with particularly heavy
movement taking place among younger people. These trends appear to
have topped out in the mid-1990s, and recent data revealed drops in the
aggregate levels of gun crime in most cities. Although the researchers did
not empirically investigate the period prior to 1993 in the YFVI cities, it is
overwhelmingly likely that the cities experienced upward movement of
gun crime levels comparable to that which occurred nationally.
Anecdotally, this is what officers in the YFVI police departments
communicated.

9. See Johnston, Patrick, "The Effects of Arrests on Gun-Related Crime in
Salinas: A Statistical Report Based on a Geometric Lag Model," in The
National Evaluation of the Youth Firearms Violence Initiative,
Cambridge, MA: Abt Associates Inc., July 1998, for a full explanation of
the geometric lag model used in the analysis.

10. It should be noted that some of the nonintensive sites, in particular
Bridgeport and Milwaukee, might well have had comparable outcome
results to those found in Inglewood and Salinas. These were not
observable in the evaluation, however, because, by agreement with COPS
and NIJ, the incident-specific data used in the intensive sites were not
collected in the nonintensive sites.

-------------------------------

Supplementary YFVI Reports

Bynum, Tim, Jennifer Frank, Athena Garrett, and Kristen Jacoby. 1998.
National Evaluation of the Youth Firearms Violence Initiative: Seattle
Case Study Final Report. Cambridge, Massachusetts: Abt Associates Inc.

Conly, Cathy, Jennifer Frank, Athena Garrett, and Kristen Jacoby. 1998.
National Evaluation of the Youth Firearms Violence Initiative: Richmond
Case Study. Final Report. Cambridge, Massachusetts: Abt Associates Inc.

Conly, Cathy, and Jack Greene. 1998. National Evaluation of the Youth
Firearms Violence Initiative: Baltimore Case Study. Final Report.
Cambridge, Massachusetts: Abt Associates Inc.

Cordner, Gary, Jennifer Frank, Athena Garrett, and Kristen Jacoby. 1998.
National Evaluation of the Youth Firearms Violence Initiative:
Birmingham Case Study. Final Report. Cambridge, Massachusetts: Abt
Associates Inc. 

Cordner, Gary, Jennifer Frank, Athena Garrett, Kristen Jacoby, Ryan
Kling, and Tom Rich. 1998. National Evaluation of the Youth Firearms
Violence Initiative: San Antonio Case Study. Final Report. Cambridge,
Massachusetts: Abt Associates Inc. 

Decker, Scott, Jennifer Frank, Athena Garrett, and Kristen Jacoby. 1998.
National Evaluation of the Youth Firearms Violence Initiative: Milwaukee
Case Study. Final Report. Cambridge, Massachusetts: Abt Associates Inc.

Decker, Scott, Jennifer Frank, Athena Garrett, Kristen Jacoby, Ryan
Kling, and Tom Rich. 1998. National Evaluation of the Youth Firearms
Violence Initiative: Inglewood Case Study. Final Report. Cambridge,
Massachusetts: Abt Associates Inc.

Dunworth, Terence, Tim Bynum, Jennifer Frank, Athena Garrett, Kristen
Jacoby, Ryan Kling, and Tom Rich. 1998. National Evaluation of the
Youth Firearms Violence Initiative: Salinas Case Study. Final Report.
Cambridge, Massachusetts: Abt Associates Inc.

Greene, Jack, Terence Dunworth, Tom Rich, Sarah Cutchins, Jennifer
Frank, Athena Garrett, Kristen Jacoby, and Ryan Kling. 1998. National
Evaluation of the Youth Firearms Violence Initiative: Final Cross-Site
Report. Cambridge, Massachusetts: Abt Associates Inc. 

Greene, Jack, Jennifer Frank, Athena Garrett, Kristen Jacoby, Ryan Kling,
Daniel McGillis, and Tom Rich. 1998. National Evaluation of the Youth
Firearms Violence Initiative: Cleveland Case Study. Final Report.
Cambridge, Massachusetts: Abt Associates Inc.

Johnston, Patrick. 1998. "The Effects of Arrests on Gun-Related Crime in
Salinas: A Statistical Report Based on a Geometric Lag Model." In The
National Evaluation of the Youth Firearms Violence Initiative.
Cambridge, Massachusetts: Abt Associates Inc.

Rich, Tom, Jennifer Frank, Athena Garrett, and Kristen Jacoby. 1998.
National Evaluation of the Youth Firearms Violence Initiative: Bridgeport
Case Study. Final Report. Cambridge, Massachusetts: Abt Associates Inc.

These supplementary reports will be available at the NIJ Web site
(http://www.ojp.usdoj.gov/nij) and the Abt Associates Web site
(http://www.abtassoc.com).

-------------------------------

Crime Classification

Procedures were developed to identify certain classes of crimes, such as
youth gun crimes. This required precise definitions of a number of terms,
as follows:

Violent crime. The standard FBI definition of violent crime--crimes
involving homicide, rape, robbery, or aggravated assault--was used in the
evaluation. 

Gun crime. For this evaluation, a gun crime was defined as any crime in
which at least one of the associated charges explicitly involved firearms.
Charges that explicitly involved firearms typically fell into one of five
categories: crimes against persons (a robbery or aggravated assault with a
firearm, for example), discharging a firearm, exhibiting a firearm,
possession of a firearm, and other firearm crimes (crimes involving illegal
sale or alteration of a firearm, for example). Some gun crimes involved a
single gun-related charge, while others involved multiple gun-related
charges (robbery with a firearm and illegal possession of a firearm by a
convicted felon, for example).

Gun arrest. As with the definition used for gun crimes, a gun arrest was
defined as one in which one of the arrest charges explicitly involved
firearms. 

Youth gun crime. A youth gun crime was defined as any crime in which at
least one of the associated charges explicitly involved firearms and at least
one person involved in the crime was a youth. Thus, the "age"
classification of a particular crime was assumed to be the age of the
youngest arrestee or suspect involved in the crime. In the earlier interim
YFVI reports and in the case study reports for the intensive sites, analyses
were presented that used different definitions of "youth," including
persons aged 14 years and under, 15 to 17 years, and 18 to 24 years. In this
Research in Brief, "youths" are defined as persons 24 years and under,
because this definition corresponds most closely to that used by the 10
police departments.

Youth gun arrest. Similarly, a youth gun arrest was any arrest in which at
least one of the arrest charges explicitly involved firearms and the arrestee
was a youth. Again, a variety of definitions of youth were used in earlier
reports; in this Brief, a "youth" is defined as a person 24 years old and
younger.

YFVI target area crimes and arrests. Mapping software, in conjunction
with the street addresses contained in the crime and arrest files, was used
to determine which crimes and arrests occurred in the YFVI target areas,
as well as in areas immediately adjacent to the target areas.

-------------------------------

Terence Dunworth, Ph.D., is managing vice president and project director
at Abt Associates Inc. This research was supported by the National
Institute of Justice and the Office of Community Oriented Policing
Services under grant number 95-IJ-CX-0069 to Abt Associates.

-------------------------------

Findings and conclusions of the research reported here are those of the
author and do not necessarily reflect the official position or policies of the
U.S. Department of Justice.

-------------------------------

The National Institute of Justice is a component of the Office of Justice
Programs, which also includes the Bureau of Justice Assistance, the
Bureau of Justice Statistics, the Office of Juvenile Justice and Delinquency
Prevention, and the Office for Victims of Crime.

-------------------------------

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

-------------------------------

NCJ 184482