Title: National Evaluation of Weed and Seed Cross-Site Analysis. Series: Research Report Author(s): Terence Dunworth, Gregory Mills, Gary Cordner, and Jack Greene Published: National Institute of Justice, July 1999 Subject(s): Drug law enforcement, drug abuse prevention and education, community policing, crime mapping, police manpower development, and problem oriented policing 86 pages 206,000 bytes ------------------------------- This is an ASCII plain-text file. To view this document in its graphic format, 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 Evaluation of Weed & Seed Cross-Site Analysis Research Report ------------------------------- U.S. Department of Justice Office of Justice Programs 810 Seventh Street N.W. Washington, DC 20531 Janet Reno Attorney General Raymond C. Fisher Associate Attorney General Laurie Robinson Assistant Attorney General Noel Brennan Deputy Assistant Attorney General Jeremy Travis Director, National Institute of Justice Office of Justice Programs World Wide Web Site http://www.ojp.usdoj.gov National Institute of Justice World Wide Web Site http://www.ojp.usdoj.gov/nij ------------------------------- National Evaluation of Weed & Seed Cross-Site Analysis Terence Dunworth Gregory Mills Gary Cordner Jack Greene July 1999 NCJ 176358 ------------------------------- National Institute of Justice Jeremy Travis Director Steven Edwards Program Monitor Prepared for the National Institute of Justice, U.S. Department of Justice, by Abt Associates Inc., under contract #95-DD-BX-0134. Points of view or opinions stated in this document are those of the author and do not necessarily represent 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. ------------------------------- Foreword Weed and Seed represents a collaborative effort of Federal, State, and local government agencies and private organizations to improve the quality of life in targeted high-crime areas of American cities. Begun in 1991, the Weed and Seed strategy appears to be deceptively simple but is, in reality, extremely complex and ambitious: drug trafficking, gang activity, and violence are carefully targeted for intervention and enforcement and community policing activities, human services programs, and neighborhood improvement initiatives are introduced to infuse new life into the community and deter future lawbreaking. As the research branch of the Department of Justice, NIJ is charged with evaluating major federally sponsored crime control initiatives. As this report makes clear, evaluating programs such as Weed and Seed poses its own set of challenges. The goal is to provide independent, objective data that communities can use to modify programs or create new approaches. At the same time, NIJ seeks to understand the nexus of crime in its community context, expanding the knowledge base in this critical area. Building on an initial evaluation of the program's implementation, NIJ then set out to assess the impact of a variety of interventions in a variety of communities with a variety of measures. Evaluators asked basic questions such as: How well did the overall Weed and Seed strategy live up to program expectations? What elements were strongest, where did they succeed or fail, and what lessons can we learn from the experience over the life of the program? Equally important, the evaluation sought to shed light on how cities and communities institutionalized their Weed and Seed programs--to what extent can communities continue the programs, interagency task forces, and organizational partnerships that were developed once official Weed and Seed funding ends? To help other communities who may want to adopt a similar strategy, evaluators asked more questions: What are the indicators of whether Weed and Seed might take root in a community? Conversely, what are the factors within a community that discourage successful implementation of Weed and Seed? Most important, the study sought to measure what impression, if any, Weed and Seed left on the program participants and community residents, the intended beneficiaries of the program. To find answers, eight cities were chosen for the national evaluation of the Weed and Seed strategy. This document synthesizes the research findings and lessons of the eight case study cities (which are available separately). As expected, the findings vary from city to city and from target site to target site (even within the same city or community). Each Weed and Seed site had its own successes. Every site had political, institutional, situational, and funding challenges to overcome. These experiences, taken together, suggest avenues for both the program and researchers to explore and ideas about how to create more successful sites in the future. These research findings provide us with reason for optimism about a comprehensive community-based approach to neighborhood safety. They will serve as guideposts to future programs designed to prevent crime and revitalize communities. Jeremy Travis Director National Institute of Justice Acknowledgments This cross-site report draws heavily on the eight companion case studies for Akron, Hartford, Las Vegas, Manatee/Sarasota, Pittsburgh, Salt Lake City, Seattle, and Shreveport. Consequently, a considerable debt is owed to those whose cooperation and support made the case studies possible. Too numerous to name here, they include U.S. Attorneys and staff, FBI personnel, city officials, and police department command and line officers. Perhaps most important were the Weed and Seed staff in each city and the neighborhood organizations whose efforts give community revitalization a fighting chance. We also appreciate the assistance of contributing staff and their skills in producing this report. Those individuals include: George Bridges, Timothy Bynum, Scott Decker, Jennifer Frank, Kristen Jacoby, Zachary Johnson, Ryan Kling, Thomas Rich, and Cheryl Roberts. Throughout the evaluation, we have benefited from the support and understanding of the two Federal agencies that we consider our partners as well as our clients. Steve Rickman and Bob Samuels, Director and Deputy Director, respectively, of the Executive Office for Weed and Seed, were unflagging in their willingness to provide information, respond to questions, critique observations, comment on reports, and generally smooth our path to the Weed and Seed sites. At NIJ, Jeff Ross and Steve Edwards were constant supporters and facilitators of our work. Of course, none of these people bear any responsibility for whatever flaws this work may have. We were able to manage that aspect of things without help. Table of Contents Executive Summary Organizational roles and community engagement Grantee organization Steering committees Building trust and community capacity Approach to law enforcement Approach to seeding Crime trends Community survey Participant interviews Conclusions Factors favoring successful implementation of Weed and Seed Community setting Program design Concentration of funds Leadership and partnership Strategic choices Note Chapter 1 Background and Objectives of Operation Weed and Seed The Weed and Seed strategy Organizational structure The U.S. Attorney's Office The Weed and Seed steering committee The weeding committee The seeding committee Weed and Seed program staff Number of sites and Federal funding Design of the National Weed and Seed evaluation Evaluation sites Akron, Ohio Hartford, Connecticut Las Vegas, Nevada Manatee and Sarasota Counties, Florida Pittsburgh, Pennsylvania Salt Lake City, Utah Seattle, Washington Shreveport, Louisiana Evaluation activities Notes Chapter 2 Program Implementation Program framework Grantees and funding levels Program goals Demonstration sites Hartford Las Vegas Manatee/Sarasota Pittsburgh Seattle Shreveport Officially recognized sites Akron Salt Lake City Organizational roles, community engagement, planning, and evaluation Role of the steering committee and other key decisionmaking bodies Public agency participation at the steering committee level Community participation at the steering committee level Role of the grantee organization Grantee institutional issues Police department Mayor's office Community-based organizations Engaging the community in building capacity Building community capacity Program planning and local evaluation Outside technical assistance and resources Data collection by the program External local evaluation Approach to weeding and community policing Multiagency task forces Prosecution Community policing and resident involvement in law enforcement and crime prevention Approach to seeding Developing a seeding strategy Seeding program partnerships Seeding funding strategies Types of programming Prevention and intervention programs for youths Neighborhood beautification Community building/community development initiatives Adult employment/economic advancement programs Family support services targeted to adults Community economic development facilitation Notes Chapter 3 Analysis of Law Enforcement Data Arrest trends in the target areas Crime trends in the target areas Possible contributing factors Crime trends in nontarget areas Preexisting crime trends in the target areas Program intensity Arrest volume Notes Chapter 4 Perceptions of Community Residents and Program Participants Distribution of respondents by age and length of neighborhood residence Personal safety and neighborhood satisfaction Severity of neighborhood crime Victimization in the neighborhood Police effectiveness and responsiveness Awareness of Weed and Seed General neighborhood quality of life General assessment Participant interviews Youth recreation and education programs Community-building programs Employment and training programs Violence prevention programs Participant suggestions for program improvement Summary overview Notes Chapter 5 Assessing the Weed and Seed Strategy Possible overstatement of program results Possible understatement of program results Factors favoring successful implementation of Weed and Seed Community setting Strength of the social and institutional infrastructure Severity of crime problems Locational advantages favoring economic development Transiency of the community population Program design Concentration of funds Funding "intensity" Channeling and leveraging other funds Leadership and partnership Notes Chapter 6 Policy Implications Program effectiveness and community mobilization Selecting sites and allocating funds Appendix 1 1995 Community Survey: Basic Interview Data Questionnaire Appendix 2 1997 Community Survey: Basic Interview Data Questionnaire Appendix 3 Significant Differences Between 1995 and 1997 in Survey Results for Nonelderly, Long-Term Residents Exhibits 1.1 Operation Weed and Seed: Number of Sites and Annual Funding 1.2 Target Area Characteristics for the Evaluation Sites 2.1Funding Designation for Weed and Seed Sites 2.2 Disposition of Hartford Weed and Seed Arrests 3.1 Part 1 Arrests Before and During Weed and Seed 3.2 Drug Arrests Before and During Weed and Seed 3.3 Part 1 Crime Before and During Weed and Seed 3.4 Change in Part 1 Crime Rate After 1 Year of Implementation: Target Areas 3.5 Change in Part 1 Crime Rate After 1 Year of Implementation: Non-Weed and Seed Areas 3.6 Change in Part 1 Crime Rate After 2 Years of Implementation: Target Areas 3.7 Change in Part 1 Crime Rate After 2 Years of Implementation: Non-Weed and Seed Areas 4.1 Distribution of Survey Respondents by Age and Length of Residence in Target Area 4.2 Survey Responses on General Neighborhood Quality of Life Evidence of Intended Changes in Measurable Crime-Related Outcomes Executive Summary Operation Weed and Seed represents an ambitious Federal, State, and local attempt to improve the quality of life in targeted high-crime areas of America's cities. First launched by the U.S. Department of Justice in 1991, the program is designed to control violent crime, drug trafficking, and drug-related crime and to provide a safe environment in which residents can live, work, and raise their families. These are, by themselves, conventional law enforcement goals. What makes Weed and Seed distinctive and innovative is that it couples community-focused human services programs and neighborhood improvement initiatives with intensified geographically targeted law enforcement efforts. Weed and Seed is administered by the Executive Office for Weed and Seed within the U.S. Department of Justice. Weed and Seed is considered a strategy, not a program. That is, Weed and Seed is a means to mobilize resources in coordinated efforts, not simply a mechanism to fund local activities that share no collective aim. The key components of this strategy are as follows: o Weeding--law enforcement efforts to remove violent offenders, drug traffickers, and other criminals from the target areas. o Seeding--human services and neighborhood revitalization efforts to prevent and deter further crime. o Community policing--proactive police-community engagement and problem solving, in which police personnel are accountable for specified geographic locations, regarded as "the bridge between weeding and seeding." The number of sites served by Weed and Seed has grown rapidly since the program's inception and currently stands at 200 sites. Eight sites participated in this national evaluation: Akron, Ohio; Hartford, Connecticut; Las Vegas, Nevada; Manatee and Sarasota Counties, Florida; Pittsburgh, Pennsylvania; Salt Lake City, Utah; Seattle, Washington; and Shreveport, Louisiana. Each of these sites had distinctive crime problems. However, they all shared high rates of violent crime related to drug trafficking and drug use. The national evaluation incorporated a wide variety of activities that included a review of funding applications and other significant program documents; individual interviews with key program administrators, senior law enforcement staff, managers of seeding activities, service providers (both current and former), and community leaders; analysis of automated, incident-level records provided by the local police departments on crimes and arrests; group interviews with participants in seeding programs; and two surveys of residents in target areas conducted in 1995 by the Institute for Social Analysis and in 1997 by Abt Associates Inc. Organizational roles and community engagement Grantee organization. The grantee and its staff assigned to the Weed and Seed effort were among the most important factors in successful program implementation at the eight evaluation sites. Sites without sufficient dedicated staff resources, from both grantees and other participating agencies, suffered in program implementation, oversight, and cohesion. Steering committees. The role and composition of steering committees varied among sites. Some steering committees were dominated by public-sector representatives, whereas others were more heavily represented and guided by community residents. For several sites, the steering committee or Weed and Seed resident bodies provided a critical means of resident participation in program decisionmaking. Building trust and community capacity. Several of the evaluation sites encountered early community resistance to Weed and Seed because residents were concerned that the exclusive focus would be on enforcement or targeted harassment. The clear lesson from these experiences was the importance of involving residents early in Weed and Seed planning, providing residents with substantial program authority, and earning their trust. The seeding program component was typically the means through which community trust was built and participation fulfilled. To implement Weed and Seed, considerable resources were devoted to effectively catalyze participation and increase capacity--even when sites had a strong preexisting community organizational infrastructure. Weed and Seed sites that employed a bottom-up, grassroots approach built trust among residents and community-based leaders and enhanced community capacity. Approach to law enforcement Law enforcement approaches across the eight sites typically included: o Increased police presence through additional personnel hours and overtime, with the majority of sites assigning dedicated officers to the target area. o Increased special operations for targeted law enforcement. o Varying degrees of increased local, State, and Federal coordination, whether in targeting offenders, narcotics operations, prosecution, or probation/parole. Local responses ranged from increased communications via monthly meetings to the creation of formal interagency and multijurisdictional task force operations housed at the same facility. o Greater concentration, coordination, and integration of efforts within police departments than before Weed and Seed and increased integration of law enforcement with seeding-type activities. o Expanded or strengthened community policing efforts or instituted new programs. o Mobilized residents who participate in crime prevention, in some cases creating effective structures for community authority and leadership. Responses ranged from increasing neighborhood watches, holding community meetings, and receiving guidance on law enforcement priorities from a citizens' advisory committee. Although weeding typically involves less resident participation than seeding, communications between residents and the police seem to have increased across sites. Overall, prosecution has been a relatively weak link in Weed and Seed due to various institutional, political, and judicial issues. In the majority of sites, there was no special Federal or local prosecution or tracking of Weed and Seed cases. In many sites, however, communication and coordination between law enforcement personnel and prosecutors improved. Approach to seeding Among the greatest challenges for Weed and Seed sites was to develop an appropriate seeding strategy with community members that effectively used Weed and Seed resources and leveraged existing community resources. Simply selecting and successfully implementing seeding programs, such as safe havens, were more difficult for most sites than implementing weeding programs, with some sites initially stumbling at this stage. Seeding was inherently a broader and more complex task, both in the development of goals and strategies and in practical organization. Seeding efforts required engaging participation and commitment from numerous types of organizations, whereas weeding had a relatively clearer mission, operating within the more established hierarchical structures of law enforcement and criminal justice organizations. Due to the broader and less defined nature of seeding, more time was needed for planning, relationship-building, and gaining consensus and commitment from the wide range of participants who shaped this domain. The eight Weed and Seed sites tried to build their programs around existing resources, in addition to creating new partnerships. Seeding program partners and providers emphasized how Weed and Seed increased coordination and communication links across neighborhood groups and other agencies. Weed and Seed sites reflected different emphases in funding local seeding programs--with varying degrees of program breadth, depth, and duration. In several sites, providing communities with authority in the seeding grant award process was critical to gaining community participation and trust. Youth programs were the primary focus of seeding activity, followed by basic neighborhood beautification efforts. Seeding initiatives undertaken by the evaluation sites can be clustered in the following order of predominance: o Prevention/intervention programs for youths, with a strong mix of programs. o Neighborhood beautification, such as community cleanups and code enforcement. o Community building and community development initiatives. o Adult employment and economic advancement programs. o Family support services aimed at adults. o Community economic development facilitation. Crime trends Across the eight sites, crime patterns varied widely. In comparing the number of Part 1 crimes in the year prior to Weed and Seed implementation with the second year of Weed and Seed, five target areas had double-digit percentage decreases (Stowe Village in Hartford, 46 percent; Crawford-Roberts in Pittsburgh, 24 percent; North Manatee, 18 percent; the Shreveport target area, 11 percent; and the Central District in Seattle, 10 percent).[1] One target area (West Las Vegas) had a single-digit decrease (6 percent), and three target areas experienced increases in Part 1 crime (South Manatee, 2 percent; Meadows Village in Las Vegas, 9 percent; and Salt Lake City, 14 percent). During this same time period, in six of nine target areas--Hartford, Pittsburgh (Crawford-Roberts), North Manatee, South Manatee, Shreveport, West Las Vegas--the Part 1 crime rates improved more than in the rest of the city or county. Also, Part 1 crimes in the Salt Lake City target area and South Manatee decreased in 1997, the latest reporting period. Although it is not possible to state definitively the extent to which different factors contributed to the observed changes in crime, a number of factors appear to have a strong correlation with these changes. For example, Hartford and Pittsburgh, which experienced the largest Part 1 crime decreases in nontarget areas, are the same two sites whose target areas achieved the largest Part 1 crime decreases. At the same time, the site with the largest Part 1 crime increase in its target area--Salt Lake City--also exhibited the largest Part 1 crime increase of all target areas. A relationship also appears to exist between crime trends and the concentration of program resources in sites that had the largest increases or decreases in crime. Hartford, for example, has the smallest target area in terms of population and area, while Salt Lake City has the largest single target area in square miles and, along with Akron, the smallest level of Federal Weed and Seed funding. Finally, changes in the drug arrest rates appear to be associated with changes in the Part 1 crime rate. For example, among the six target areas for which arrest data are available, the four reporting decreases in Part 1 crime from the year prior to Weed and Seed through the second year of implementation (Hartford, Pittsburgh, North Manatee, and Shreveport) all experienced initial high rates of drug arrests--suggesting an initial period of intense weeding activities--followed by declining drug arrest rates. Assuming the level of enforcement as measured by police presence has remained somewhat constant, this trend reflects success in reducing drug activity. The Salt Lake City and South Manatee target areas both experienced large increases in the number of drug arrests in 1997 compared to 1996, suggesting that perhaps these sites had not yet succeeded in reducing the level of drug activity in the target areas. Thus, across these six sites, the changes in drug arrest volumes follow the same general pattern as the changes in Part 1 crimes. Community survey To understand the perspective of community residents, a survey of target area residents was conducted for all eight sites at two separate time intervals--in 1995 and 1997. The objective of the survey data collection was to measure changes in residents' perceptions of public safety, crime, police performance, general neighborhood quality of life, and awareness of the Weed and Seed program. To the extent possible, the same methodology was used in both surveys. There were, however, some notable differences. The 1995 survey consisted of inperson interviews, while the 1997 interviews were conducted by telephone. In addition, the 1997 survey consisted of fewer questions than the 1995 survey. The decision to proceed in 1997 with telephone interviewing and a shortened instrument was based on the difficulties experienced in 1995 in completing the targeted number of interviews per site. The methodological differences between the two surveys, combined with underlying population changes in the evaluation sites, led to a shift between 1995 and 1997 in the demographic profile of respondents in each surveyed target area. A decision was made to focus the survey analysis on a demographically comparable subset of respondents: nonelderly, long-term residents. The sites appear to fall into three groups with respect to the overall pattern of survey findings comparing 1995 and 1997 among nonelderly, long-term residents. o Manatee and Pittsburgh exhibited substantial evidence of changes in residents' perceptions across multiple outcome measures, including severity of crime and police effectiveness in controlling crime. o Akron, Hartford, and Seattle exhibited some evidence of changes in residents' perceptions of selected crime dimensions, either drug-related crime (Akron and Seattle) or violent and gang-related crime (Hartford). o Las Vegas, Salt Lake City, and Shreveport exhibited little evidence of changes in residents' perceptions of general public safety or the severity of specific types of crime in the neighborhood. Participant interviews To gain the perspective of community residents whom the seeding programs were intended to benefit, interviews were conducted with seeding program participants. These interviews were not intended to measure outcomes of the programs that were visited; rather, they were designed to learn the perceived benefits and drawbacks of the programs from the individuals who participated in them. Long-term effects of these programs on the lives of the program participants cannot be deduced from the interviews. The seeding programs that were visited fall into three general categories: (1) youth recreation and education; (2) employment and training; and (3) violence prevention. Based on participants' comments, the seeding programs appeared to provide services that otherwise would not have been available in the target areas. Most of those interviewed also indicated that participation in the seeding programs was a positive experience that helped them feel more secure emotionally, physically, or both. The general themes that emerged focused on providing additional structure and discipline in the lives of target area youths and opportunities and assistance for adults to work toward personal and professional growth. Conclusions The experience of the eight participating Weed and Seed sites raised two questions: Are sites achieving the intended changes in measurable outcomes? What factors appear to promote successful implementation of the program and thus promote achievement of the intended results? Based on the pattern of findings with respect to the rate of Part 1 crimes and crime-related survey questions, one can group the sites into four categories, first according to the evidence of reduced Part 1 crimes and then (within each category) according to the evidence of improved public perceptions: o Hartford and Pittsburgh showed strong evidence of reduced Part 1 crimes and improved public perception of crime-related measures. o Manatee/Sarasota (North Manatee) and Shreveport exhibited substantial evidence of reduced Part 1 crimes. o Seattle, Akron, Las Vegas (West Las Vegas), and Manatee/Sarasota (South Manatee) exhibited some evidence of reduced Part 1 crimes. o Salt Lake City and Las Vegas (Meadows Village) showed no evidence of reduced Part 1 crimes. Those sites in the first group (Hartford and Pittsburgh) stand out among the rest, with strong evidence of favorable outcomes with respect to both crime and public perceptions of crime. Factors favoring successful implementation of Weed and Seed What factors appear to have promoted successful implementation of the program and achievement of the program's intended results? To address this question, one needs to consider the site characteristics and program features that appear to distinguish the target areas as categorized above, according to their measurable crime-related outcomes. Community setting. There are preexisting features of the program setting that may make Weed and Seed easier or more difficult to operate effectively. Important factors included the strength of the social and institutional infrastructure, the severity of crime problems, locational advantages favoring economic development, and transiency of the community population. Program design. The mix of weeding and seeding activities and the sequence of component implementation appear to be important factors in gaining community support for the program. Important factors included early seeding, sustained weeding, high-level task forces combined with community policing, and an active prosecutorial role. Concentration of funds. Sites appeared to have greater success if they concentrated their program resources on smaller populations, especially if they could similarly channel other public funds and leverage private funds. The important factors included funding intensity along with channeling and leveraging other funds. Leadership and partnership. Finally, a less tangible ingredient that seemed to characterize the more successful programs was the active and constructive leadership of key individuals. By its very nature, Weed and Seed places a great premium on effective coordination among groups with different organizational missions, responding to different constituencies. To establish effective working relationships among these organizations required personal energy and initiative. The most effective implementation strategies were those that relied on bottom-up participatory decisionmaking approaches, especially when combined with efforts to build capacity and partnership among local organizations. This required a longer term perspective about the program and its potential to bring about community change. Such sites--including some that achieved substantial crime reductions within the time period analyzed here--have established a stronger foundation and more sustainable basis for further community-targeted initiatives. Strategic choices The strategic choices now faced by policymakers in charting the future direction of Weed and Seed revolve around issues of designating sites for continued funding, selecting sites for new awards, and allocating funds among participating sites over time. The experience of the eight sites studied in the this evaluation suggests that Weed and Seed has affected the target areas through either (or both) of two avenues. The first, termed program effectiveness here, relates to the specific initiatives that focus on law enforcement and crime prevention. The second, called community mobilization here, is the process in which Weed and Seed provides a catalyst for greater involvement of neighborhood residents and community-based organizations. In the interest of program effectiveness and successful community mobilization, the following inferences can be drawn from site experiences: o Weed and Seed should seek the highest feasible concentration of funds in the program sites. Given the annual funding constraint provided by the congressional appropriation, this implies a more selective process in choosing sites to receive new awards and/or some shortening in the number of years that ongoing sites receive program funding. o Weed and Seed should place its funding priority on sites with geographically small target areas and with favorable community settings and programmatic designs when selecting new program sites--i.e., with features favoring successful implementation, as described in chapter 5. o Weed and Seed should consider additional effort in providing technical assistance to the funded sites, given the importance of institutional capacity-building and infrastructure development. The lessons learned from the more successful sites--and the less successful ones--on these issues can be generalized to a large degree, and new sites should receive the benefit of this experience. This seems especially important in promoting the partnership arrangements that characterize the more successful programs examined in this research. Note 1. Part 1 crimes include homicide, rape, robbery, aggravated assault, burglary, larceny, and auto theft. Chapter 1 The U.S. Department of Justice launched Operation Weed and Seed in 1991 as an initiative to control violent and drug-related crime and to make communities safe for residents and businesses. The premise was simple: use law enforcement tactics and resources to "weed" criminal activities from targeted high-crime areas and "seed" those neighborhoods with human services and revitalization projects to prevent and deter further crime. This distinctive collaboration between law enforcement and human services relied on effective community policing--officers who not only removed criminals from the streets, but who also engaged residents in crime deterrence efforts and encouraged their participation in "seeding" programs. The organization with which this was accomplished varied among the eight sites evaluated. Background and Objectives of Operation Weed and Seed Operation Weed and Seed represents an ambitious attempt to improve the quality of life in targeted high-crime areas of America's cities. First launched by the U.S. Department of Justice in 1991, the program is designed to control violent crime, drug trafficking, and drug-related crime and to provide a safe environment in which residents can live, work, and raise their families. These are, by themselves, conventional law enforcement goals. What makes Weed and Seed distinctive and innovative is the means by which these goals are achieved--by coupling community-focused human services programs and neighborhood improvement initiatives with intensified, geographically targeted law enforcement efforts. Weed and Seed is administered by the Executive Office for Weed and Seed (EOWS) within the U.S. Department of Justice (DOJ). The program now operates in 200 sites. Current annual funding (for fiscal year 1999) is $49 million, including asset forfeiture money. The Federal oversight responsibility for each participating site rests with the U.S. Attorney's Office for the corresponding district. This decentralized arrangement (versus one orchestrated from Washington, D.C.) was intended to reinforce local control while providing a more hands-on Federal role--in particular, to enable Federal prosecutors to be more responsive to local law enforcement initiatives. The Weed and Seed strategy Those who federally administer Operation Weed and Seed consider it a strategy, not a program. That is, Weed and Seed is a means of mobilizing resources in coordinated efforts, not simply a mechanism to fund local activities that share no collective aim. The key components of this strategy are as follows: o Weeding--law enforcement efforts to remove violent offenders, drug traffickers, and other criminals from the target areas. o Seeding--human services and neighborhood revitalization efforts to prevent and deter further crime. o Community policing--proactive police-community engagement and problem solving, with police personnel accountable for specified geographic locations, regarded as "the bridge between weeding and seeding."[1] Each of these three strategic components was viewed as a necessary element for program success. Program effectiveness also required close coordination among all three. EOWS characterizes Weed and Seed as an "incubator for social change" to stabilize the conditions in high-crime communities and thus promote community restoration.[2] In brief, the process by which the program seeks to effect such change is as follows: o Additional resources for law enforcement and concentrated efforts of police and prosecutors on targeted high-crime areas can more effectively identify, arrest, and prosecute criminals, especially those engaging in narcotics trafficking and violent crime. o More effective crime detection and response, speedier investigations and trials, and stricter sentences serve to put drug dealers and street criminals out of action. Additionally, others may be discouraged from criminal behavior. o Awareness of the results of heightened law enforcement among community residents and businesses means they will more readily cooperate with community-assigned police and prosecutors, further promoting arrests and prosecutions. By gaining the trust and support of the community, police and prosecutors engage residents and businesses as problem-solving partners in the law enforcement effort. o Human services programs, especially ones aimed at youths, address the school, family, or peer group risk factors associated with criminal behavior. Afterschool, weekend, or summer youth activities; adult literacy classes; parental counseling; and similar programs improve the conditions that might otherwise lead to the use of drugs or violence. o Reduced incidence of crime and enhanced perceptions of public safety, along with improved housing, stronger schools and other public services, and heightened community involvement in neighborhood beautification improve the climate for economic development. Organizational structure The coupling of human services with heightened law enforcement distinguishes Weed and Seed from other location-specific crime-reduction strategies. The program's local organizational structure is distinctive and has the following features: o The U.S. Attorney's Office provides Federal oversight and coordinates Federal, State, and local law enforcement and prosecutorial activities, as well as general DOJ oversight of the Weed and Seed strategy. o The Weed and Seed steering committee establishes operational goals, designs and develops programs, guides implementation, and assesses program achievements. The steering committee is headed by the U.S. Attorney and includes the mayor (or elected county official); the district attorney; the chief of police; other appropriate elected or appointed Federal, State, or local officials (who control the allocation of resources potentially available to serve the program's objectives); private-sector representatives; and target area residents. The steering committee is further organized into separate weeding and seeding components. o The weeding committee plans and monitors the law enforcement efforts, including interdiction and prosecution. o The seeding committee plans and monitors the prevention, intervention, treatment, and neighborhood restoration efforts. o Weed and Seed program staff are responsible for operating program activities. Normally, a Weed and Seed program director has overall management responsibility. Day-to-day program activities are operated by a weed coordinator (usually a supervisory member of the local police department) and a seed coordinator. Another organizational element found in virtually all Weed and Seed sites is the safe haven, defined by EOWS as "a multiservice center where a variety of youth and adult services are coordinated in a highly visible, accessible facility that is secure against crime and illegal drug activity" and "a place where youths and other residents can access needed services, develop relationships, find opportunities to be productive and successful, and enhance skills."[3] Indeed, EOWS requires all new Weed and Seed sites to have at least one safe haven.[4] Number of sites and Federal funding The number of sites served by Weed and Seed has grown rapidly since the program's inception. Initial grants were awarded in fiscal year (FY) 1991 to Kansas City, Missouri; Trenton, New Jersey; and Omaha, Nebraska; another 17 sites were added in FY 1992, 1 in FY 1993, 15 in FY 1994, 52 in FY 1996, 30 in FY 1997, 58 in FY 1998, and 24 in FY 1999. None were added in FY 1995. Until very recently, funded sites were either officially recognized sites or demonstration sites. Officially recognized sites qualified for initial program funding in annual amounts of less than $250,000 by having previously implemented Weed and Seed-like initiatives in their designated target areas. Demonstration sites qualified for full program funding in the range of $500,000 to $750,000. Currently, most funded sites receive $225,000 annually. Total program funding for Weed and Seed rose rapidly in its early years, from $11.5 million in FY 1992 to $31.5 million in FY 1994. Over the succeeding 4 years, funding increased further by more than one-third, reaching $42.5 million in FY 1998 and $49 million in FY 1999, as shown in exhibit 1.1. (The exhibit includes funds provided to the program from several sources--primarily from the Weed and Seed Program Fund, but also from the Byrne Discretionary Grant Fund and the Asset Forfeiture Fund.) Design of the national Weed and Seed evaluation This report is the first multisite evaluation of Weed and Seed to include findings on both the implementation of the program and measurable outcomes related to crime and public safety. As early as 1993, single-site local evaluations have been conducted in many participating sites. (Starting in 1994, EOWS required sites to use some of their funding to support local evaluations.) In addition, the implementation of the program in the first 19 sites to receive Weed and Seed funding was studied under the National Process Evaluation of Operation Weed and Seed.[5] However, this earlier evaluation did not analyze any data on crime rates or perceptions of public safety. Evaluation sites Eight sites participated in this evaluation as identified in exhibit 1.2. These sites were selected by the U.S. Department of Justice as examples of different aspects of Weed and Seed. For each site, the evaluation focused on one or two Weed and Seed target areas, as follows:[6] Akron, Ohio: The West Side neighborhood has the largest population-- nearly 24,000--among the target areas included in the evaluation, with a mix of both renter- and owner-occupied housing (one-third of which is considered substandard) and several retail districts. Unlike the other participating areas, the area's crime rate is somewhat below the corresponding citywide average. This site was unfunded at the time of its designation for the national impact evaluation. Hartford, Connecticut: Stowe Village, one of the city's most impoverished neighborhoods, is a public housing development in its northeast neighborhood. This densely populated, multiracial community consists of 550 housing units in 31 two- or three-story buildings. This site was selected for funding on the basis of the 1994 competitive solicitation. Las Vegas, Nevada: Meadows Village is a small and shrinking community at the north end of the tourist district. It is populated primarily by transient, foreign-born, primarily Spanish-speaking residents who make or seek their livelihood in low-skilled service jobs related to the city's entertainment industry. This site was also selected for funding on the basis of the 1994 competitive solicitation. West Las Vegas, located further north and separated from the downtown area by a major interstate highway, is a larger and less transient area than Meadows Village. It has a predominantly black population, an active commercial district, and established social infrastructure. This site was added in 1995. Manatee and Sarasota Counties, Florida: This is a multijurisdictional site, selected for funding through the 1994 competitive solicitation and for participation in the Weed and Seed National Performance Review project. Weed and Seed was implemented in six local target areas, two of which were studied in the impact evaluation. North Manatee has a predominantly minority population; more than three-fourths of the residents are black. More than one-third of the residents are elderly, and nearly half of the adult residents did not graduate from high school. One-fourth of the residents receive public assistance. South Manatee consists primarily of low-income duplex rental housing. Compared to North Manatee, this area has a younger and more transient population with a smaller percentage of minority and elderly residents. Nearly one-half of the households contain children, many with single parents. Pittsburgh, Pennsylvania: The Hill District is located between the city's main business center and its educational/cultural district. Historically a center for black culture, commerce, and entertainment, this community has experienced a dramatic population decline during recent decades. The area consists of six separate neighborhoods, some experiencing unemployment rates approaching 50 percent. This site was selected for funding in 1992 and for participation in the Weed and Seed National Performance Review project in 1994. Salt Lake City, Utah: The West Side includes three neighborhoods (Glendale, Poplar Grove, and a large section of Fairpark) that together make up a larger geographical area (6.3 square miles) than any of the other evaluation sites. Housing conditions and socioeconomic characteristics vary greatly across these communities. Overall, most of the housing is owner-occupied, but some areas have high concentrations of vacant or boarded buildings. Some neighborhoods are relatively stable, working-class areas; others are more transient and poverty-ridden. This site was selected for funding and participation in the National Performance Review in 1994. Seattle, Washington: The Central District is a highly congested section east of downtown, consisting of eight separate neighborhoods with both residential and retail areas. The diverse population includes both university students and low-income families, with poverty rates in some neighborhoods approaching 50 percent. This site was initially selected for funding in 1992. Shreveport, Louisiana: The Highland-Stoner Hill area, the least densely populated of the evaluation sites, is one of the city's oldest neighborhoods. Economic difficulties in the oil and gas industry have contributed to a decline in the community's business sector, declining property values, and a deterioration in housing. This site was selected for funding in the 1994 competitive solicitation, and it also participated in the Weed and Seed National Performance Review project. Although each of these areas had distinctive crime problems, they all shared high violent crime rates related to drug trafficking and drug use. Most also had serious gang-related crime problems. These sites were selected for the national evaluation during the fall of 1994. Pittsburgh and Seattle were chosen from among the first 19 Weed and Seed demonstration sites, which received their initial demonstration funding in FY 1992. Hartford, Las Vegas, Manatee/Sarasota, and Shreveport were chosen from among the sites to receive their first demonstration funding in FY 1994. Akron and Salt Lake City were officially recognized sites in FY 1995 (and then subsequently became demonstration sites). Evaluation activities The national evaluation incorporated a wide variety of evaluation activities at each site that included a review of funding applications and other significant program documents; individual interviews with key program administrators, senior staff within law enforcement agencies, managers of seeding activities, service providers (both current and former), and community leaders; analysis of automated, incident-level records provided by the local police departments on crimes and arrests; group interviews with participants in seeding programs; and two surveys of residents in target areas, conducted in 1995 and in 1997. Based on these activities, a separate case study was first prepared for each of the eight evaluation sites. There are numerous challenges to evaluating a comprehensive, locally targeted initiative such as Weed and Seed. In the end, it is impossible to reach any definitive conclusion about Weed and Seed's effectiveness; what is most important, there is no reliable basis on which to determine what would have happened in these sites in the absence of Weed and Seed. Nonetheless, it is possible to examine carefully the manner in which the program was implemented and to consider the changes in measurable outcomes that occurred following implementation. One cannot attribute such changes to Weed and Seed alone, but one can draw from the experience of these sites to suggest those factors that appear to have favored successful implementation of the Weed and Seed strategy and that appear to have promoted achievement of the intended results. Notes 1. EOWS regards community policing as "increasing police visibility and developing cooperative relationships between the police and citizenry in the target areas." The associated techniques include foot patrols, police ministations, nuisance abatement, victim referrals to support services, and community relations activities, in which the community is encouraged to undertake such initiatives as neighborhood watches, citizen marches and rallies, drug-free zones, and graffiti removal. See Executive Office for Weed and Seed, Operation Weed and Seed Implementation Manual, Washington, D.C.: U.S. Department of Justice, Office of Justice Programs, Executive Office for Weed and Seed: 1-3, 1-4. 2. Ibid., 9-6. 3. Executive Office for Weed and Seed, Weed and Seed Fiscal Year 1998 Program Guide and Application Kit for New Sites, Attachment 2, Washington, D.C.: U.S. Department of Justice, Office of Justice Programs, Executive Office for Weed and Seed, 1998: 2-1. 4. In its literature, EOWS describes safe havens as an integral part of a "risk factor and protective factor" approach to crime prevention. This is viewed as the counterpart to comprehensive community-based disease prevention programs that have operated effectively in the public health arena--for example, in combating heart disease. In the Weed and Seed context, risk factors, those that make one susceptible to criminal behavior, must be identified and addressed. At the same time, protective factors, those that serve to counter or neutralize risks, must be enhanced. 5. Roehl, Janice A., et al., National Process Evaluation of Operation Weed and Seed, Research in Brief, Washington, D.C.: U.S. Department of Justice, National Institute of Justice, 1996, NCJ 16124. 6. Several of these eight sites, including Las Vegas and Pittsburgh, have implemented Weed and Seed in additional target areas not mentioned here. However, because this additional experience has been relatively recent, we have focused this report on the target areas noted, for which substantial postimplementation data have been collected. ------------------------------- Chapter 2 The most successful Weed and Seed efforts required the full cooperation of human services programs, community buy-in, and, perhaps most important, systemic commitment beginning with police officers on the street and extending through the areas' political structure and prosecutors' offices. The initiatives also relied on strong seeding programs, including prevention and intervention programs for youths, neighborhood beautification, community building and development efforts, adult employment and economic advancement programs, family support services, and general community economic development. Together, these elements were to achieve three key program objectives: o Coordinate, concentrate, and integrate public and private resources in target areas. o Empower residents to solve neighborhood problems. o Increase investment and commitment from the private sector. Program Implementation The Weed and Seed strategy had three key objectives: o Coordinating, concentrating, and integrating public and private resources in target areas. o Empowering residents to assist in solving neighborhood problems. o Increasing private-sector investment and commitment. Although the eight sites successfully implemented elements from the weeding, community policing, and seeding program components, each site had distinct local contexts which shaped its approach and community responses.[1] Program framework Grantees and funding levels. Exhibit 2.1 shows Weed and Seed grantee organizations and funding awards for the eight evaluation sites. Grantees included mayors' offices, local police departments, and local nonprofit 501(c)(3) organizations. Among the demonstration sites, the Pittsburgh program began first, in April 1992, followed by Seattle in January 1993. The remaining demonstration sites were awarded their first grants in Federal fiscal year 1994, implementing their programs in late 1994 or early 1995. Salt Lake City and Akron became officially recognized sites in FY 1995 but did not receive significant funding until FY 1996 and FY 1997, respectively. As a result of the timing of its funding, the Akron Weed and Seed program operated as an officially recognized but unfunded site at the time of the evaluation. Among the demonstration sites, program duration and target area size and population, all of which affect the breadth and intensity of the Weed and Seed intervention and implementation strategies, need to be considered. At the two extremes, Hartford received about the same amount of funding for 1,300 people in a 0.1-square-mile target area as did Manatee/Sarasota for 33,641 people in 6 noncontiguous target areas spanning more than 40 square miles. Program goals. Having been selected to participate in Weed and Seed, the eight sites shared the fundamental goals of eliminating violent crime, drug trafficking, and drug-related crime and providing a safe environment for residents. The sites also set goals according to the other major Weed and Seed components which included drug abuse prevention, especially activities for youths; expansion of community policing efforts, including strengthening community-police relations and increasing resident and business owner participation in crime prevention; and neighborhood restoration, such as code enforcement, improving housing stock, and attracting new investment. The following summarizes some of the variations in local goals. Demonstration sites Hartford. Because the Stowe Village target area was a city center for narcotics and gang activity, improving public safety was the site's primary goal. Program goals included reducing youth violence, narcotics trafficking, and violent crime; reducing fear and reenfranchising youths into the community; and returning community control to law-abiding citizens. Key objectives included supplementing the U.S. Department of Housing and Urban Development (HUD) Family Investment Center and other existing social services with youth and family support programs, prevention activities, and employment-related training, as well as mobilizing residents to participate in crime prevention and seeding activities. Las Vegas. Suppressing high-level drug crime and youth gangs were key goals for the two Las Vegas target areas, with the objectives of: (1) targeting and eliminating the 50 most disruptive and violent offenders, and (2) increasing community support for law enforcement. Seeding goals included increasing child and family health care, strengthening youth development and resiliency, and improving housing stock and property appearance. In the Meadows Village target area, which had a more transient population and a relatively high proportion of immigrants, objectives included reducing legal and language barriers of residents, developing a shared computer-based information system in the safe haven to enhance the availability and effectiveness of health care, education, and human services; and establishing a neighborhood association to promote the identification and resolution of common community problems. Manatee/Sarasota. Law enforcement goals were to reduce street-level drug dealing (and the open-air drug markets), reduce violent and property crimes, increase citizen participation in neighborhood action teams, and implement and integrate a community policing program in Manatee County, where one previously did not exist. The seeding focus was on youth-oriented or crime prevention programs, including education, job training, and recreation, as well as the establishment of safe havens. A central seeding component of the Manatee/Sarasota program was community empowerment, including leadership development and providing the community authority over nonenforcement domains of the program. Pittsburgh. A key weeding goal was to eliminate open-air drug trafficking, whereas seeding goals included integrating and expanding economic development strategies, providing employment and job training opportunities, and improving housing conditions. Pittsburgh's program focused on building stronger community-level organizations to enable communities to address their problems by obtaining resources through community and economic development. The key strategies employed to achieve the operational goals were: (1) providing training to community members and Weed and Seed participating organizations, and (2) systematically increasing the community's access to information. Seattle. In addition to controlling street-level gang-related drug and violent crimes, goals included developing an effective managerial approach to neighborhood problem solving involving community and business groups, extending the partnership concepts to include agencies at the State and Federal levels, and mobilizing residents to work with police in problem solving. Seattle focused on youth antiviolence enforcement and youth-oriented or crime prevention projects, as well as supporting broader community efforts promoting employment (skills training and business development), education, and health care. A collective decisionmaking process among community groups, residents, and participating public agencies provided a broad foundation for the program. Shreveport. In addition to reducing drug trafficking, law enforcement goals included reducing weapons violations, property crimes, and gang influence; targeting and removing repeat offenders; increasing community education on police services, crime prevention, and drug awareness; and increasing participation in and effectiveness of the neighborhood watches. Prevention goals centered on youth risk factors and education, and neighborhood restoration focused on vigorous code enforcement with resident participation. A key seeding emphasis was establishing safe havens and afterschool programs (which were sorely lacking in the target area), as well as coordinating existing public and private services and resources to make them available to residents. Officially recognized sites Akron. This city received little funding during most of the national evaluation period. Although Akron had not yet implemented its program, strategies had been developed. Law enforcement goals included eliminating drug markets and crack houses and reducing burglaries and prostitution. Special community policing goals included restoring neighborhood pride and achieving a higher percentage of high school graduates. Seeding goals built on an existing program, the Crouse Caring Communities Project. Goals included developing a model for integrated service delivery to children and families, working with the city's public schools to increase achievement and decrease counterproductive behavior, building employment skills of residents, improving housing, and further developing the business district. Salt Lake City. In its first year as an officially recognized site, the city's goal was to extend technical assistance for community mobilization activities in the three target neighborhoods and to empower the community by providing small grants to fund high-priority projects for residents. With increased recognition-level funding, law enforcement goals included reducing levels of violent crime related to gang and drug activities, focusing on high-level drug cases for Federal prosecution, and increasing the implementation of community policing efforts, including problem solving with the community and the integration of community- oriented police officers with beat officers. Seeding goals included increasing the coordination and planning around economic development and addressing risk/protective factors for youths and families, including intensive case management and community assessment. Salt Lake City was already a recipient of a major Comprehensive Communities Program grant and intended to use Weed and Seed to build on those efforts. Organizational roles, community engagement, planning, and evaluation Among the eight sites, there were substantial variations in the role of the steering committee, the role of the grantee organization, how the community was engaged in capacity building, and the role of planning and evaluation in the implementation of the Weed and Seed program. Role of the steering committee and other key decisionmaking bodies In general, the steering committees played an important role in: (1) establishing goals and objectives, (2) providing guidance and oversight on key program design and implementation issues, and (3) integrating weeding with seeding at the policy level. At most sites, the steering committee also played a critical role in coordination across agencies, sectors, and jurisdictions. Across all eight sites, steering committee members included key public agency representatives and local government officials; variations occurred around inclusion of social service providers, community-based organizations, and residents. Although some steering committees were dominated by public-sector agencies, others were more heavily composed of and guided by community residents. Public agency participation at the steering committee level. Participation and commitment among the public-sector steering committee members varied considerably by site. The following illustrates the range of public officials' involvement on the steering committee. At one site, a central problem was a lack of participation by key agency officials, who sent delegates who had no decisionmaking authority to committee meetings and did not bring to bear the resources of their agencies on behalf of Weed and Seed. In this program, the mayor, U.S. Attorney, and district attorney had relatively limited program involvement. As one might imagine, with a lack of prosecutorial commitment at the highest levels, prosecution remained a serious problem for this site, with no Federal prosecutions and no special State attention for Weed and Seed cases. At this same site, however, community participation and leadership in the steering committee were strong, as were the leadership and capabilities of the Weed and Seed program staff. These two factors enabled successful implementation of many program elements, in spite of the lack of participation of relevant agency heads. Conversely, another Weed and Seed site demonstrated too much involvement and ownership by agency heads, such that conflict developed between grantees and other agencies over the direction and implementation of the program. The U.S. Attorney at this site was highly involved in most aspects of the program but did not share the same vision for the program as the local grantee organization. This conflict between the two key leaders caused considerable disruption in the management of the program and seems to have inhibited local ownership of the program.[2] Although the role of the U.S. Attorney varied, the U.S. Attorney's Office in a majority of the evaluation sites struck a balance between the two more extreme cases described above. (See the section, "Approach to weeding and community policing," in this chapter, for a discussion of Federal prosecution issues.) Community participation at the steering committee level. For some sites, the steering committee and Weed and Seed neighborhood bodies provided the primary means of formal community participation in program decisionmaking. Other sites incorporated resident participation at the subcommittee level for weeding and seeding. The following illustrates the range of community participation in key Weed and Seed decisionmaking bodies: Seattle set up two primary governing bodies: (1) the steering committee, which included the relevant agency heads with some social service organizations, and (2) the Citizens Advisory Committee, consisting of residents and community organizations. The steering committee provided policy-level guidance and program oversight, while the Citizens Advisory Committee provided program direction for both weeding and seeding activities. A collective decisionmaking process among community groups, residents, and participating public agencies provided a broad foundation for the program. In Manatee/Sarasota Counties, the neighborhood action teams had substantial program authority, guiding most of the nonenforcement efforts in the target areas. There were six neighborhood action teams--one for each target area--consisting of residents and representatives of organizations, businesses, and law enforcement operating in the target area. In addition, there was one overall steering committee for seeding and one for weeding; the seeding committee comprised elected representatives from the different neighborhood action teams, whereas the weeding committee comprised law enforcement representatives. Shreveport provides an example where community participation in decisionmaking was achieved primarily through its steering committee. The steering committee included representatives of relevant public agencies, but residents and community and nongovernmental organizations accounted for more than half of the membership. The program relied on its steering committee for policy-level guidance and oversight, with program staff providing day-to-day management of weeding and seeding program components.[3] In Las Vegas, the steering committee--with representatives from law enforcement and social service groups--functioned as the key decisionmaking body for the first 2 years of the program. Later, a smaller executive committee made up of executives or designees of some key agencies was established as a first-line management structure. In spite of a strong executive committee, decisionmaking remained decentralized for weeding and seeding, with residents helping to guide seeding and community policing activities.[4] In Hartford, the steering committee predominantly comprised public agency officials and service providers. Attempts were made to foster greater participation from the tenants' association and residents, but the tenants' association was relatively inactive, and there was a lack of community-based organizations in Stowe Village. Steering committee meetings were, however, open to the public. The steering committee typically made strategic decisions, although program staff were more responsible for program management and implementation. The Seattle and Manatee/Sarasota program organizations show a more bottom-up approach, with residents granted high decisionmaking authority through formal resident bodies. Hartford, on the other hand, lies closer to the other end of the spectrum, where public agency representatives were primary decisionmakers. The remaining six sites fell somewhere in the middle of this spectrum. Role of the grantee organization The grantee and its staff assigned to the Weed and Seed effort were probably the single most important factors in successful program implementation among the evaluation sites. The capabilities, vision, and commitment of Weed and Seed program staff and the grantee institution were more important factors in program implementation than the type of grantee organization. To implement the Weed and Seed strategy, it was necessary for sites to have sufficient dedicated staff resources to guide the program, engage resident participation and ongoing community leadership, build the necessary network of relationships in the public and private sectors, oversee implementation, and provide technical assistance to the communities. Sites without sufficient dedicated staff resources, from both grantees and other participating agencies, suffered in program implementation, oversight, and cohesion. Structuring and setting in motion a comprehensive and coordinated strategy was the principal challenge for sites. Effective program staff leadership enabled implementation success in spite of serious obstacles, and staff problems (including inadequate staff, lack of competencies for the given role, and misunderstanding of the mission) severely impeded program implementation. Even in such cases, however, some program elements succeeded due to the leadership of other Weed and Seed participants or to changes in staff. In keeping with the concept of Weed and Seed as a strategy, the organization, staffing, and method of implementation evolved over time at different sites. Weed and Seed was not a static program structure, and lessons were often incorporated as the process unfolded. Shreveport, for example, initially had inexperienced and insufficient staff, which led to a long delay in implementing seeding programs and resulted in the failure of the first afterschool program associated with Weed and Seed. One problem was that the seed coordinator was a volunteer whose full-time job was to run a large afterschool program. This person was not prepared for these dual roles, nor were there appropriate Weed and Seed management and accountability structures in place. Eventually a full-time seed coordinator was hired, and the Weed and Seed program director was replaced. With strong staff leadership and community participation, implementation occurred smoothly, with considerable accountability. Consistent with the philosophy of building community capacity, Weed and Seed programs also had to find the right balance between providing needed assistance to communities and challenging them to mobilize more community resources. The Manatee/Sarasota Weed and Seed effort reorganized its initial program structure to reduce the number of staff involved with neighborhood coordination. Although neighborhood coordinators were initially helpful in mobilizing residents, too much staff assistance decreased community initiative and was not an effective use of resources. Grantee institutional issues The Weed and Seed evaluation sites demonstrated how different grantee institutions can bring different institutional strengths, weaknesses, and opportunities to the program and thus affect its design and operations. Although the capabilities of the Weed and Seed program staff and leaders are more important than the type of grantee institution, the latter is a key factor in structuring a balanced program. The following discusses these potential strengths and weaknesses. Police department. Since public safety and law enforcement are central to Weed and Seed, it is important to encourage a high level of commitment from police departments. Having a police department as grantee can facilitate the engagement of greater police department resources and ongoing department commitment. However, a police department as grantee can also create distinct disadvantages. Police departments, for example, are less equipped to handle the social services programming and community consensus-building process than other types of grantees (given the more hierarchical structure of police departments). For example, the Shreveport police department, as grantee, initially encountered difficulties of this nature but later adapted well. Furthermore, placing Weed and Seed in a police department also tends to identify the program more strongly as a law enforcement initiative, at least until seeding is implemented. In Seattle, where the community policing department was the grantee, Weed and Seed tried to overcome this perception by giving the seeding component to the city Department of Housing and Human Services. This arrangement probably facilitated community acceptance and broader integration of seeding initiatives in the city.[5] Given this concern about a focus on enforcement, it is interesting to note in sites where the police department was the grantee that there was no evidence of disproportionate focus and spending on law enforcement efforts. In fact, Seattle and Hartford spent far more of their budgets on seeding programs than on weeding. Mayor's office. Importantly, the mayor's office can bring the strength of a city's full resources to the program and can translate program successes in the target area to the city at large. The mayor as grantee can also facilitate governmental capacity building and broadly strengthen collaboration with citizens and community organizations. Pittsburgh provides an example of these strengths. At one site, having a mayor as grantee placed the program within a volatile political environment; the program became dominated by individual political concerns and turf wars. Further, changes in administration reduced program commitment in one city, while at another it signaled the beginning of a more focused and productive effort.[6] A mayor's office, however, brings both the assets and liabilities of political office, which vary by city and shift over time. Community-based organizations. Among the eight evaluation sites, only Manatee/Sarasota Counties had a nonprofit organization as grantee. The grantee was Drug Free Communities, Inc., a nonprofit organization created by the mayor. Although it was unusual for a Weed and Seed grantee to be anything other than a police department, mayor's office, or similar government agency, this approach seems to have worked well for Manatee/Sarasota. As a practical matter, because the Manatee/Sarasota Weed and Seed program stretched over two counties, it would have been difficult for any single local government agency to serve as effectively as the grant-receiving entity. In addition, Drug Free Communities already had established strong grassroots organizational and political support in Manatee County and had received a significant Federal grant from the Center for Substance Abuse and Prevention (CSAP). Although based on limited data, the Manatee/Sarasota experience suggests other community-based organizations might effectively serve as future Weed and Seed grantees and can bring distinct strengths, including community trust. It would be important, however, for such organizations to have had a broad-based community constituency and positive experience working with local governmental organizations, in addition to a demonstrated capacity to manage Federal grants.[7] Engaging the community in building capacity Two central tasks of the Weed and Seed strategy are to engage residents in problem solving from a comprehensive view and to gain commitment from community members and key partners to construct and implement an integrated strategy. Several evaluation sites encountered early community resistance to Weed and Seed due to resident concerns of exclusive focus on enforcement or concerns of targeted harassment. o When the city of Seattle proposed applying for funding from the Weed and Seed program in the spring of 1992, many community groups expressed public opposition, fearing the proposed weeding programs would be measures to harass and control persons living in the target area. After heated public debate, the mayor organized and held meetings with community coalitions and representatives from the public agencies involved with the proposal. These meetings culminated in revisions to the city's proposal which increased the operational role of the Citizens Advisory Committee, composed of residents and community organizations within the Weed and Seed area. o In the first 2 years of the Las Vegas Weed and Seed program, there was negative publicity and organized opposition over concern that its enforcement efforts would target young minority males. Eventually, as the seeding component gained momentum and genuinely involved grassroots community groups, these concerns were mitigated. o In Pittsburgh, in the first 2 years of Weed and Seed, the program focused primarily on enforcement and rarely involved residents in program decisionmaking. Residents had expected to see community development initiatives, not just crackdowns, and had become alienated and angry about perceived broken promises. Subsequently, the program was radically restructured to focus on community capacity building, with substantial seeding program authority given to residents. According to leaders of community organizations and residents, Weed and Seed sites that employed a bottom-up grassroots approach built trust among residents and community organizations. As one community leader said, "There's a trust level that had been developed with local people.... This gives [the program] a chance to mushroom." The clear lessons from these experiences were the importance of involving residents early in Weed and Seed planning, providing residents with substantial program authority, and earning their trust. The seeding program component was typically the means through which community trust was built and participation fulfilled. Citizen willingness to participate in law enforcement was sometimes contingent on the trust built and demonstrated through the investments in seeding. Building community capacity. Preexisting community organizational infrastructures varied across target areas. In sites such as Meadows Village in Las Vegas, a lack of social infrastructure made program implementation difficult. Even when sites had strong community organizational infrastructures, considerable resources were needed to effectively catalyze participation and increase capacity to implement Weed and Seed. Pittsburgh and Manatee/Sarasota, for example, focused on community capacity building and devoted considerable ongoing resources to these efforts. The following are examples of positive effects of such investment: o One Manatee community-based leader emphasized the importance of Weed and Seed's work with the neighborhood action teams to "put programs into action" that met community needs. He said, "I feel good about what I'm doing because we [the community] do this for ourselves." He said that for too long, the approach was to go after a quick fix and do what benefited the provider more than the community. Outsiders were brought in to "fix them up," only to take their funds and leave the community worse off. o Another community leader said, "We didn't bring anyone into the community ... We did it ourselves in the community not with outside services. This gives empowerment to the community; it shows our youths the good example of seeing leaders active in their own community." Weed and Seed also provided a forum for leadership development among residents. In Salt Lake City, for example, a resident took the lead in the initial Weed and Seed grant application and played an instrumental role in early program implementation. In the North and South Manatee target neighborhoods, there emerged charismatic individual leaders who essentially ran the safe havens as volunteers and directed a variety of community activities. For all sites, it remained a central organizational challenge to continuously develop and broaden leadership to implement and sustain the Weed and Seed strategy. Program planning and local evaluation Sites employed a variety of methods to improve their strategic planning, including outside technical assistance, data collection by the program, and external local evaluations. In general, sites would have benefited from more technical assistance early in the Weed and Seed planning process. Some demonstration sites felt they were not sufficiently prepared when they received the grant money and would have benefited from a more rigorous planning process, including technical assistance and hiring a local evaluator, before receiving the full grant award.[8] Weed and Seed goals and strategies needed to be focused--especially on the seeding side--with clear criteria as to how resources should be allocated and progress measured.[9] Sometimes this was achieved, sometimes not. Strategic focus also needed to be integrated with other goals, such as community participation and capacity building. Outside technical assistance and resources. Several sites engaged outside technical assistance or resources to aid program planning. The following are illustrations of some initiatives: o In the second year of the Shreveport program, the Communities That Care strategic planning process provided a clear focus for Weed and Seed planning and programming, especially on the seeding side. A total of 70 people participated in a process that identified several key community issues and risk factors. Subsequent subcontract requests for proposals (RFPs) and the local evaluation focused on these issues and risk factors. Although this process proved to be time-consuming, it built consensus and strengthened program coherence for both seeding and community-oriented policing. Prior to that, there were no established criteria from which to select seeding programs and on which to build consensus among community groups represented on the steering committee. Consensus building was particularly important for this site because the target area encompassed two diverse neighborhoods in which community-based organizations did not share the same visions of community change; otherwise, program efforts may have been more diffuse and fragmented. o Manatee/Sarasota Counties also benefited from the Communities That Care process, except technical assistance came rather late in the program, and community members thought the process too academic. Manatee also employed a process of "asset mapping" the communities' strengths that was used to build upon the Weed and Seed efforts.[10] This process drew from the work of John Kretzmann and John McKnight of Northwestern University's Neighborhood Innovations Network.[11] o In Pittsburgh, community organizations received training under the Communities That Care program. The Pittsburgh program also received technical assistance from the National Congress on Community Economic Development to assess the status of community economic development and to provide training to communities. o Salt Lake City used its National Performance Review Laboratory grant to refine local planning to improve the city's chances for being selected as a Weed and Seed site, as well as to aid communities to clearly identify local resources that might be drawn into crime reduction planning and community improvement efforts.[12] Assessing community strengths and needs with its members led to effective use of resources and helped to build program coherence and consensus. Appropriate technical assistance can help grantees with methodology and provide an objective outside perspective; this may be particularly important for grantees lacking experience in this kind of planning and consensus-building process. A partially funded 6-month to 1-year planning period might be beneficial for some sites. Data collection by the program. Some sites undertook systematic data collection efforts as part of their ongoing program planning and management and made such information available to target area residents. In Manatee/Sarasota Counties, for example, Weed and Seed staff conducted community surveys and maintained a demographic database on the target areas, which groups accessed for grant applications. The site also maintained a community resource library and provided ongoing technical assistance to a variety of community groups, with an average of 15 requests per month. External local evaluation. Half of the national evaluation sites--Shreveport, Seattle, Las Vegas, and Hartford--undertook external local evaluations. In three of the sites, the local evaluation played a constructive role in program assessment. Shreveport Weed and Seed provides an example in which the local evaluation played a central role in program planning and assessment. Officials there felt that having a local evaluator was a critical factor in successful program implementation. The firm they hired had experience in evaluating comprehensive community initiatives and built a collaborative relationship with program staff by using quarterly evaluation information to guide program improvement.[13] Shreveport officials' major regret was that the program was nearly 2 years old before a local evaluator was hired, and they believe better decisions could have been made--especially on the seeding side--if evaluation information had been available sooner.[14] Approach to weeding and community policing Most sites developed and implemented coherent law enforcement strategies that responded to local conditions and incorporated stronger street-level patrols with some degree of higher level interagency coordination. Weeding approaches across the eight sites typically included: o Increased police presence through additional officer hours and overtime, with the majority of sites assigning officers dedicated to the target area. o Increased special operations for targeted law enforcement, especially for drug-related and violent crime. o Greater concentration, coordination, and integration of efforts within police departments than before Weed and Seed. o Varying degrees of increased local, State, and Federal communication and coordination, whether in targeting offenders, narcotics operations, prosecution, or probation or parole. The assignment of dedicated officers to the Weed and Seed target areas was important in building relationships with residents and in aiding enforcement through better knowledge of the neighborhood, better intelligence, and the ability to operate proactively. In Shreveport, for example, many residents knew the four dedicated "bumble bees," the bicycle patrol officers in yellow shirts and black shorts; and, in Pittsburgh, residents successfully lobbied to keep one of their Weed and Seed officers from being reassigned. Most sites executed intensive initial weeding efforts, focusing on drug enforcement and special joint operations.[15] At many sites, such as Hartford and Las Vegas, drug enforcement efforts targeted major offenders; other sites, including Pittsburgh and Manatee, focused more on street-level dealers. The strategy depended in part on preexisting law enforcement operations and, in some cases, concerns of residents. Pittsburgh and Manatee already had other operations under way that focused on the major dealers, and the weed task forces collaborated with these efforts. Drug enforcement and suppression tactics typically included use of informants, undercover operations (such as buy-and-bust tactics and reverse sting operations), execution of search warrants, saturation patrols, and surveillance. The Shreveport weed officers also employed a "zero tolerance" approach, sometimes setting up checkpoints in problem areas. In addition to drug-related crime, property crimes were a major concern in Shreveport and Manatee; early weeding efforts concentrated on improving this situation.[16] Both sites have also focused more on quality-of-life issues, especially as crime levels have been reduced.[17] Increased interagency collaboration was a central component of most weeding strategies. Such efforts included monthly meetings between weed officers and probation/parole officers to catch parole violators, frequent joint Federal and local special narcotics operations, and the creation of formal interagency Weed and Seed law enforcement task forces, like those in Las Vegas, Manatee/Sarasota Counties, and Pittsburgh.[18] Multiagency task forces Although multiagency task forces concentrated on the target area, they pursued drug cases across jurisdictional lines. Consequently, the benefits of these efforts extended beyond the target areas, particularly when the focus was on high-level drug dealers who controlled large operations. Task force missions varied according to the nature of crime in the target areas and preexisting law enforcement operations, so Weed and Seed task forces complemented existing efforts. Examples of task forces include: o In Las Vegas, the Joint Las Vegas Metropolitan Police Department-FBI (LVMPD-FBI) Task Force focused on relatively high-level drug distribution in the West Las Vegas target area. The U.S. Attorney and FBI organized the task force and helped foster Federal commitment to the program. In an approximately 2-year period, the task force was responsible for 137 arrests for State crimes and 40 Federal arrests and prosecutions. o In Manatee/Sarasota Counties, the cornerstone of the weeding strategy was the creation of a multijurisdictional interagency Violent Crimes Task Force (VCTF), which focused on street-level drug dealing and violent crimes in the six target areas.[19] Coordinated by the FBI, VCTF included the four local law enforcement agencies, the Bureau of Alcohol, Tobacco and Firearms (ATF), the U.S. Marshal's Service, and the Florida Department of Law Enforcement. The task force involved the full-time commitment of personnel by all participating agencies--Federal, State, and local--and the officers all worked from a central location.[20] Between October 1994 and March 1997, the task force made 2,766 arrests in both counties, including 54 Federal cases. In the North and South Manatee target areas, VCTF produced 231 cases during this period.[21] o The Pittsburgh Weed and Seed task force, coordinated by the narcotics division of the Pittsburgh Police Department, focused on street-level narcotics traffic in the target areas, while working collaboratively with an established multijurisdictional task force that focused on major violent drug traffickers in the Pittsburgh area. Prosecution Overall, prosecution has been a relatively weak link in Weed and Seed due to various institutional, political, and judicial issues. Although interagency communications often increased as a result of Weed and Seed, in the majority of sites, there was no special Federal or local prosecution or tracking of Weed and Seed cases.[22] Shreveport provides an example of a site with no Federal prosecution and no special efforts for local prosecution or sentencing; there, program officials and residents were extremely dissatisfied with the "revolving door syndrome," with repeat offenders often released on parole.[23] Manatee and Sarasota Counties, on the other hand, appear to be one of the few sites in which prosecution and cooperation with Weed and Seed has been vigorous at both the Federal and local levels. At the local level, although the district attorneys typically did not allocate special resources to Weed and Seed, several sites specially stamped cases, and communications often increased and improved between law enforcement personnel and prosecutors. Although there was usually no formal preference given to Weed and Seed cases ("a serious case is a serious case," as one State attorney's office representative said), law enforcement officials felt that such communication improved the focus on important Weed and Seed cases. In Las Vegas, prior to 1996, Weed and Seed cases did not receive special "fast track" attention in the prosecutor's office. However, since that time, all Weed and Seed task force cases have been screened and presented to the district attorney by task force personnel. This has resulted in fewer of these cases being dropped. As an illustration of varying local philosophies, the district attorney at a Weed and Seed site with strong prosecution felt the biggest benefit of Weed and Seed was its ability to concentrate resources in the worst areas and to focus on quality-of-life crimes. Conversely, a district attorney at a site with extremely poor prosecution felt, politically, that his office could not give any more attention--show favoritism--to one geographic area over another. In general, the political environment in which district attorneys operate, as well as limited resources, constrained special local prosecution of Weed and Seed cases. At the Federal level, the U.S. Attorney's participation and Federal prosecution was also uneven. Shreveport and Seattle processed no Federal prosecutions for Weed and Seed cases, and Akron Weed and Seed cases were prosecuted at the local and State levels. In Pittsburgh, Weed and Seed participants initially attempted to take more cases federally, but there was resistance from Federal judges. In Shreveport, where drug cases typically did not have enough weight to qualify for Federal guidelines, the U.S. Attorney's Office was unwilling to lower Federal guidelines to make the case.[24] However, for drug cases that did not meet normal Federal prosecution standards in Manatee and Sarasota Counties, the U.S. Attorney sometimes lowered the Federal drug-level requirements to pursue Weed and Seed cases.[25] In Las Vegas, Weed and Seed task force cases received strong Federal prosecution. The Hartford Weed and Seed program, which had great success in arrests of major gang narcotics traffickers, illustrates some criminal justice challenges. As illustrated in exhibit 2.2, as of June 1997, more than half of the Hartford Weed and Seed's 612 State-level cases were dismissed, with only 18 percent of arrests resulting in incarceration. For felony arrests which resulted in incarceration, 39 percent received a sentence of less than 2 years; the disposition distribution for felony drug arrests was nearly identical to that for all felony arrests. Hartford did have important Federal-level prosecutions of key gang leaders, however, resulting in strong sentencing. One factor to consider is that Weed and Seed provides substantial resources for law enforcement to generate more arrests, but the criminal justice component often did not receive additional resources to pay for any additional staffing or reorganization. Prosecutorial cooperation depended primarily on politics, preexisting institutional capacities, and individual dispositions. In several cases, Weed and Seed increased prosecutorial cooperation--through increased communication or leadership of the U.S. Attorney at the Federal level. Nonetheless, the criminal justice component generally did not occur as envisioned in the original Weed and Seed concept.[26] Community policing and resident involvement in law enforcement and crime prevention Weed and Seed enabled most sites to expand or strengthen community policing efforts or to institute new programs. Although some sites, such as Seattle, already had well-established, proactive community policing programs, others had difficulty integrating community policing with traditional law enforcement to adopt a problem-solving approach; in a majority of sites, Weed and Seed advanced the process of integration. Manatee County, for example, had a highly traditional police department and had not yet embraced community policing methods; participation in Weed and Seed enabled the county to launch its first community-oriented policing program.[27] As they got to know the neighborhoods, the community police officers also became key informants for the Weed and Seed Violent Crimes Task Force. Community policing initiatives at various sites, including nuisance abatement, landlord programs, graffiti eradication, code enforcement, and neighborhood cleanups, helped improve property maintenance and neighborhood environments. Officers also participated in a wide range of youth recreation and education programs in the target areas; such activities engaged local youths in constructive activity, provided positive role models for youths, and built community relations. In addition to enhancing community policing efforts, Weed and Seed provided a vehicle for mobilizing residents to participate in crime prevention, in some cases creating effective structures for community authority and leadership. In Seattle, the Weed and Seed Citizens Advisory Committee determined weeding priorities; and in Manatee County, where community policing was new, the Weed and Seed Neighborhood Action Teams provided a structure for citizens to work with community policing officers. For many sites, neighborhood watches became an increasingly important means of citizen participation in crime prevention. Weed and Seed led to more neighborhood watches and invigorated participation at many sites; watch participants reported high satisfaction with their participation and police support. At these sites, Weed and Seed seems to have improved police enthusiasm for neighborhood watches, as well. Although weeding has typically involved less resident participation than has seeding, communications between residents and the police increased across sites through enhanced community policing efforts, public training and education projects, regular neighborhood meetings, and increased police participation in seeding-related programs. Law enforcement personnel reported getting more information from residents and increased responsiveness and cooperation in enforcement and crime prevention. The nature of resident involvement in law enforcement also was influenced by preexisting organizational structures and cultures in both the police departments and communities. For example, in Seattle, where the Citizens Advisory Committee exerted authority over weeding, there was a strong culture of citizen participation and a long history of community policing. A key strength of the Salt Lake City Weed and Seed program was its effective mobilization of residents in law enforcement and crime prevention. Through the Comprehensive Communities Program, Salt Lake City had already established an extensive network of mobile neighborhood watches, with more than 800 participating residents and community action teams that engaged in neighborhood problem solving. Weed and Seed built on and strengthened this network. Such organizational infrastructure facilitated citizen participation in crime prevention and law enforcement. For a majority of sites, police substations in central community locations provided focal points for police contact with the community and furthered program integration. For example, Hartford created a substation in Stowe Village, and Salt Lake City, Seattle, and Manatee created substations at their safe havens. Having a central location for the weeding and seeding also helped some sites integrate the program and provided a focus for community engagement. Shreveport, for example, had a dedicated Weed and Seed site, which housed the weeding and seeding staff and served as a community center and focal point for community participation and response. In Manatee/Sarasota, the grantee, Drug Free Communities, was co-located at the Police Athletic League with the Weed and Seed multiagency Violent Crimes Task Force. Although Weed and Seed generally seemed to increase resident involvement and improve community-police relations, achieving a partnership with community authority is an ongoing and more difficult process than selective participation. Such participation and communications, however, are critical steps in building that partnership. Weed and Seed advanced the dialogue significantly. Approach to seeding Seeding typically was the gateway to community participation and engendered the most community support. In many sites, providing communities with authority in the seeding grant award process was critical to gaining community participation and trust. In Seattle and Manatee/Sarasota, neighborhood bodies (the Weed and Seed Citizens Advisory Committee and the Neighborhood Action Groups) chose programs to fund. An executive of a national youth program in Manatee/Sarasota said Weed and Seed makes it clear to providers that the neighborhood makes the decisions, and the programs must justify themselves to the neighborhood. She said Weed and Seed tries to involve communities to accomplish community-determined needs. She described it as a grassroots program, where Weed and Seed staff listen to what the community says and try to figure out how to make it happen. Pittsburgh took the approach of giving community boards authority over a substantial portion (35-40 percent) of the seeding budget, with the community body responsible for issuing RFPs, making program selections, and managing contracts and projects. Shreveport more broadly involved the community in the initial planning phase via a public strategic planning process to determine the seeding strategy and establish community participation on the steering committee responsible for approving seeding programs. Shreveport issued RFPs, with staff making initial program recommendations to the committee. At many sites, the RFP process provided valuable experience for participating community-based organizations and for neighborhood review committees. Las Vegas used the RFP process as an opportunity to train potential seeding providers through a 2-day workshop in grant preparation and management. Pittsburgh and Manatee/Sarasota also provided technical assistance in grant writing to community-based organizations. In Hartford, no RFPs were issued because the steering committee generally knew which programs it wanted to fund. In their own ways, sites had to balance issues related to strategic coherence of the program with goals of providing authority to residents and engaging their participation in solving neighborhood problems; this was where staff and community leadership played critical roles. Developing a seeding strategy Among the greatest challenges for Weed and Seed sites was developing an appropriate seeding strategy with community members to most effectively target and leverage existing community resources. Simply selecting and successfully implementing seeding programs, such as safe havens, in itself was more difficult for most sites than implementing their weeding programs. Seeding was inherently a broader and more complex task, both in the development of goals and strategies and from a practical organizational standpoint. Seeding efforts required securing participation and commitment from numerous types of organizations, whereas weeding had a relatively clearer mission, operating within more established hierarchical structures of law enforcement and criminal justice organizations.[28] Due to the broader and less defined nature of seeding, relatively more time was needed for planning, relationship building, and gaining consensus and commitment from the wide range of participants who shape this domain. It is difficult to compare and assess seeding strategies across sites because they arose from local context--the existing resources and programs, capacities, cultures, and needs. Since sites varied widely in contexts and preferences, the purpose of the seeding effort was to unite people in a process that would build momentum toward neighborhood restoration, creating a multiplier effect over time. Implementation during Weed and Seed was only the beginning of this process, which has a longer time horizon. Seeding program partnerships The eight Weed and Seed sites built their programs around existing resources, in addition to creating new partnerships. For example, Hartford Weed and Seed integrated seeding efforts with the existing Department of Housing and Urban Development Family Investment Center, and the Salt Lake City Weed and Seed program built on the Comprehensive Communities Program. Extensive partnerships and collaborations were created through Weed and Seed in the public and private sectors, with most of the private-sector partnerships with nonprofit and community-based organizations. In Shreveport, for example, Weed and Seed arranged for public services to come to the target area, including free immunizations for children from the State department of public health, a library bookmobile, and free computer training from a local university. In Seattle, the Washington Insurance Council joined Weed and Seed to establish the Seattle Neighborhood Action Program as a private/public partnership to help revitalize a neighborhood.[29] Seeding program partners and providers emphasized how Weed and Seed increased coordination and communication links between neighborhood groups and other agencies. Almost all respondents said they felt more connected to the community and service providers. A director of a youth service organization in Shreveport said, "Establishing partnerships is the key to Weed and Seed--cooperation on projects and working together to bring the community together. [You] can touch more people as a group of organizations than as a single entity." Many providers said working with Weed and Seed enhanced their organizational capacity and enabled them to receive other grants to continue their efforts. There were some former providers, however, who said they accomplished their goals with Weed and Seed funding but, when that funding ended, they were unable to sustain the same level of services.[30] While it is too early in implementation to know which program elements will be sustained, these observations reinforce the importance of using Weed and Seed to build community partnerships and capacity rather than as a primary means to fund services. Seeding funding strategies Sites reflected their different priorities by funding seeding programs with varying degrees of breadth, depth, and duration. For example, Shreveport concentrated on funding five afterschool and summer youth programs that previously did not exist in the target area, combining substantial reach with intensity. These programs will receive about 2 years of Weed and Seed funding before providers have to find replacement funding. (It appears they may be successful through a combination of city and private resources.) Conversely, Seattle initially funded more than 20 different programs, mostly related to prevention and education activities for youths and employment training, spreading funding more thinly across a variety of organizations. In a later stage of the program, funding was concentrated on a few key programs. Similarly, Salt Lake City initially provided partial grants to numerous organizations but later chose a smaller number of core programs to support. Manatee/Sarasota, which also emphasized youth programs, focused on program sustainability by providing only partial funding for programs and requiring programs to develop partnerships and other funding sources.[31] In the two Manatee target areas, funding allowed new community-based organizations to form, others to expand existing programs, and innovative programs to be tested. In the Hill District, Pittsburgh funded a broader range of projects, using money to establish and strengthen community organizations but avoid creating competing structures. Training also has been a central component of Pittsburgh's strategy, teaching and assisting communities to develop greater self-reliance. In addition to funding a strong mix of programs for youths, the Hartford Stowe Village program provided a range of family support services. Types of programming Seeding initiatives undertaken by the evaluation sites can be clustered as follows, according to predominance: Prevention and intervention programs for youths. Sites focused most heavily on prevention-oriented programs, with a strong mix of afterschool programming and safe havens, recreation and sports, skills and employment training, job development, and health/substance abuse-related programs. By concentrating on this program dimension, many sites achieved both significant reach and intensity among this segment of residents. Neighborhood beautification. Sites also incorporated neighborhood beautification efforts, such as neighborhood cleanups and code enforcement. Shreveport used Weed and Seed funds to pay for a full-time city code enforcer for the target area; this site also organized numerous neighborhood cleanup efforts, which relied upon community groups, the Shrevecorps service youths, and prison community-service workers. South Manatee had a strong community leader who led code enforcement and cleanup efforts. Residents and program staff expressed great satisfaction with results of such beautification efforts. Community building/community development initiatives. Initiatives specifically aimed at community development were underused in some programs, given the importance of this dimension in the Weed and Seed strategy. Of course, community development is both a process objective and a desired outcome. Furthermore, community building occurs as a byproduct of working together on common goals, such as various community groups undertaking a neighborhood cleanup initiative. Nonetheless, some sites developed initiatives that had community capacity building as the central goal. Pittsburgh, in particular, made this a key program strategy, funding or organizing a variety of programs, which included: o Workshops on community development through the Community Technical Assistance Center. o Neighborhood technology centers. Weed and Seed organized an effort to create and maintain a community computer network, bringing together partners that had not previously worked together, including public schools, universities, libraries, neighborhood organizations, and the city. Recently, the police department provided the community network with Internet access to its new crime data mapping system. o A community development reference manual. With Weed and Seed and other funds, the Center for the Community Interest developed a manual to assist Pittsburgh community organizations in addressing neighborhood problems. Hartford hired a professional community organizer for Stowe Village, but that effort did not succeed in mobilizing residents as intended. A resident building captain program is now being implemented to increase resident capacity for self-governance. Across the sites, the safe havens and community centers supported by Weed and Seed provided a focal point for residents and community organizations. Regular outreach by Weed and Seed staff, including safe haven coordinators at some sites, played a vital role in increasing community awareness and gaining participation in neighborhood activities. For example, Shreveport Weed and Seed published a monthly calendar of events and programs in the target area. Adult employment/economic advancement programs. Hartford, Seattle, and Pittsburgh undertook training initiatives for adults, as well as youths. Computer programs were often well-received by the neighborhoods, including Seattle's Mid-town Commons Advanced Computer Training Project, which provided residents with Internet training and access. Hartford Weed and Seed installed a computer lab, which became very popular among Stowe Village residents who received computer and educational training and Internet access. Family support services targeted at adults. Hartford employed more family support service interventions than most sites because approximately 95 percent of its resident population was unemployed. Similarly, Las Vegas undertook relatively more family support initiatives, particularly in the Meadows Village target area, which had a relatively transient population and a high concentration of recent immigrants. Community economic development facilitation. Although staff at some Weed and Seed sites participated in economic development initiatives, this program component usually received the least attention, particularly in the beginning, when more basic neighborhood concerns needed to be addressed. Pittsburgh was one of the few sites that focused on economic development in any systematic way. In Pittsburgh's Hill District, located near the business center, the Weed and Seed umbrella brought various parties together around economic development issues. The Hill District had already received a major urban revitalization grant, and community organizations were active in economic development planning. Weed and Seed facilitated and advanced these efforts. Conversely, economic development prospects were far different in Pittsburgh's Hazelwood target area--a former mill town--which was less geographically desirable for investors. Economic development, part of the neighborhood restoration phase of Weed and Seed, was a program area that was difficult to define for most sites. Of course, the principal goal of Weed and Seed was to improve conditions to make economic development possible. One needs to consider the short timeframe covered in this discussion of program implementation--only 2 to 4 years--when economic development is the program component least under the control of a small target area and most dependent on a variety of prerequisites. The implementation of such Weed and Seed elements, as discussed above, provided target neighborhoods with positive steps in the direction of community restoration. To assist in this implementation, EOWS provided technical assistance to a number of sites. Notes 1. Please refer to the case studies for a more indepth review of each site's program. 2. The steering committee has an inherently complicated management structure because the U.S. Attorney is its designated head, yet the grantee is the organization responsible for fiscal management and day-to-day program operation. The Operation Weed and Seed Implementation Manual suggests that the primary role of the U.S. Attorney, as leader of the steering committee, is to bring Federal resources to the program-- particularly in the area of criminal justice--and to use the authority of that office to advance consensual program goals. 3. Shreveport also held and publicized monthly neighborhood Weed and Seed meetings. Weed and Seed program officials found initially that their steering committee membership was overly large (30 members), and participants did not understand the magnitude of the commitment involved. Eventually, however, a core group of 18 that met monthly operated more effectively. 4. Five seed committees were formed--grants, community empowerment, planning and strategies, youth focus, and budget administration. 5. Conversely, dividing Weed and Seed across city departments could potentially increase program fragmentation and lead to a lack of integration of weeding and seeding, without coordination. Having a program in one place creates a synergy that otherwise may be harder to achieve, depending upon preexisting relationships and program leadership. 6. Such political changes affected not only the role of the mayor's office but also other participants including the U.S. Attorney. 7. On the other hand, potential weaknesses of community-based organizations include lack of resources (as compared with a city grantee) and less power to institutionalize changes citywide. With less institutional authority, community-based organizations must rely more on consensus building. 8. In addition to strategic planning, sites also must prepare for Federal grant management and the required fiscal reporting; at least two cities had difficulties setting up the necessary accounting systems. 9. In a site such as Hartford's Stowe Village, where Weed and Seed funding per capita was greatest, a variety of programs could be funded with significant reach and intensity while funding lasted; it is unclear, however, how this will be sustained. For sites with larger populations and multiple neighborhoods, careful planning and targeting became even more important. 10. The Manatee/Sarasota program continues to advance this process as it makes the transition from a demonstration to a training site. 11. Kretzmann, John, and John McKnight, Building Communities from the Inside Out: A Path Toward Finding and Mobilizing Community Assets, Evanston, IL: Northwestern University, Center for Urban Affairs and Policy Research, Neighborhood Innovation Network, 1993. 12. Pittsburgh Weed and Seed used NPRL funds to hire a consultant to assess the training needs of the community-based organization in the first and second Weed and Seed locations. The program also used the funds to automate financial recordkeeping and to enhance the technology project that provided community residents with greater access to information. Program participants had anticipated the NPRL funds would enhance the city's access to other Federal funds--this leveraging did not occur because the participation of other Federal agencies in this effort was minimal. The Shreveport Weed and Seed program decided to decline its NPRL grant award after the site was unable to get clear Federal NPRL program guidelines. 13. Incorporating the risk factors strategic approach into the evaluation criteria, the evaluation firm relied upon a community survey, interviews with project participants, and program documentation. Evaluation staff also helped Weed and Seed staff review and assess proposed new seeding programs. 14. In Shreveport, the first evaluation firm was dismissed because it could not recognize and diagnose significant initial implementation problems; the evaluator painted a false, rosy picture instead of providing constructive analysis and feedback. 15. Shreveport did not begin with a major crackdown but rather with the permanent assignment of a small number of additional officers to the target area. 16. In its preweeding effort, Manatee initially focused on reducing property crime. 17. For example, Shreveport has recently focused on such issues as truancy, public drinking, and loitering. 18. Salt Lake City Weed and Seed worked closely with the existing Metro Narcotics Task Force that focused on high-level drug dealers. 19. The Manatee/Sarasota program decided the most appropriate strategy for the area was to target six noncontiguous neighborhoods simultaneously. Law enforcement officials believed they could have a much greater impact through a simultaneous, coordinated effort so that crime would not simply be displaced from one neighborhood to the other. Displacement problems would otherwise have been exacerbated by having four different local law enforcement agencies that did not coordinate their activities prior to Weed and Seed. Additional factors in having the six target areas included the desire to rely on greater leveraging of resources and issues of fairness. 20. While the agencies funded the core positions, a second law enforcement position and enforcement overtime were funded by Weed and Seed; Weed and Seed also paid for Violent Crimes Task Force overhead, including rent, equipment, and training. 21. Violent Crimes Task Force report, June 1997. 22. Exceptions include the Las Vegas Weed and Seed task force and the Hartford Police Department, which track Weed and Seed cases. 23. Typically, cases would get plea bargained with a sentence of probation, often with probation on top of probation for repeat violators. The Weed and Seed steering committee, not including the U.S. Attorney, eventually wrote letters to judges expressing the community's concern about the probation problem; the letters were received positively by the majority of judges, and several judges met with the steering committee to discuss this and related issues. In response to this sentencing problem, Weed and Seed officers also met monthly with State probation and parole officers to catch people in violation of parole, which leads to mandatory sentencing. 24. Other complicating factors for sites such as Shreveport included inadequate court capacity and judicial dispositions toward sentencing that were beyond the control of Weed and Seed. 25. Manatee/Sarasota Counties already had strong prosecution and sentencing prior to Weed and Seed; cooperation increased for Weed and Seed cases, however. 26. As outlined in the Operation Weed and Seed Implementation Manual. 27. Weed and Seed led to Manatee County's COPS grant award, which paid for its community policing staff. 28. Of course, with community-oriented policing, the boundaries expand more. 29. This was one of the few examples of a substantial business-sector partnership established in the Weed and Seed sites. 30. One provider indicated a year of funding was not enough time to integrate a new service into her program. 31. Program staff also bartered services. ------------------------------ Chapter 3 Throughout the program, Weed and Seed officials targeted criminal activity within the target areas, focusing specifically on seven Part 1 crimes--homicide, rape, robbery, aggravated assault, burglary, larceny, and auto theft. The reduction of these crimes was generally viewed as one of the most important Weed and Seed goals and a key indicator of its success. Results were generally positive. However, there was some variation by site according to preexisting crime trends and the concentration of program resources. Analysis of Law Enforcement Data To analyze arrest and crime patterns in the Weed and Seed target areas, researchers requested computerized incident-level data from law enforcement agencies in each of the eight Weed and Seed sites, including basic facts about each arrest and crime (for example, when and where the crime was committed) and each criminal charge associated with each arrest and crime. Because the primary interest was in arrests that were made and crimes that occurred in the target areas, procedures were developed to identify those arrests and crimes in the jurisdictionwide files. In most cases, the procedures involved geocoding the address where the arrest was made or where the crime occurred.[1] The focus was also on the seven Part 1 crimes. Although these crimes account for only one segment of the overall crime problem (ignoring, for example, order maintenance and other crimes closely associated with quality-of-life issues), there are standard definitions of these crimes that all law enforcement agencies follow, thus allowing for site-to-site comparisons. Given that controlling drug trafficking and drug-related crime is one of the key Weed and Seed objectives, researchers also focused on drug arrests.[2] Finally, researchers analyzed annual changes in these arrests and crimes. More detailed analyses focusing on monthly trends are contained in the case study reports on each site. Arrest trends in the target areas Chapter 2 discussed how the evaluation sites implemented the weed component of their Weed and Seed strategy. In general, the approaches included some form of increased law enforcement presence in the target areas. One indicator of the degree of success in weeding is the number of arrests made, which is examined in this section.[3] Exhibits 3.1 and 3.2 show, respectively, annual figures for arrests for Part 1 crimes and for drug arrests.[4] "Weed and Seed years," rather than calendar years, are shown in the exhibits. In both exhibits, the annual total number of arrests, the annual number of arrests per 1,000 residents, and the percentage change in arrests from the preceding year are shown for both the Weed and Seed target area and for the remainder of the jurisdiction. With the exception of Salt Lake City, only complete (12-month) years are shown.[5] Salt Lake City is presented differently because only 5 months of arrest data were available prior to the start of its Weed and Seed program in August 1995.[6] Thus, to have any basis for comparison to the period prior to Weed and Seed, year-to-year comparisons in Salt Lake City were based only on data from March through July. Annual arrest rates are therefore not shown for Salt Lake City. Figures for Las Vegas and Akron also are not shown in these exhibits--felony arrest data from Las Vegas were not available, and less than 2 complete years of arrest data from Akron were provided. The annual arrest rates illustrate some differences across the sites and between individual target areas and the remainder of each site's corresponding jurisdiction. Hartford consistently has the highest Part 1 and drug arrest rates per 1,000 persons across the evaluation sites. Hartford's rates, in particular, are about 10-15 times higher than drug arrest rates in any of the other sites. Ignoring Hartford, the Part 1 arrest rates are fairly similar, although Pittsburgh's and Seattle's drug arrest rates are about twice those in the other sites. In comparing the target areas to the rest of their jurisdictions, in four of the six sites (North Manatee, South Manatee, Seattle, and Shreveport), the target area Part 1 arrest rate is comparable to the rate in the remainder of the jurisdiction. All of the target areas have significantly higher drug arrest rates than their surrounding corresponding jurisdictions. In interpreting year-to-year changes in arrest rates, one might expect to find high arrest rates during the first year or so of Weed and Seed implementation, reflecting an initial intense level of weeding. If these initial weeding activities were successful, then, as implementation continues into later years, one might expect a decline in the number of arrests, assuming there was a fairly constant level of enforcement in the target area. Exhibits 3.1 and 3.2 show this occurred in most of the Weed and Seed sites. For example, drug arrest rates declined in five of the seven target areas--Hartford, Pittsburgh, North Manatee, and, to a lesser extent, Seattle and Shreveport. Drug arrest rates in both Salt Lake City and South Manatee had not yet shown any decline. Changes in Pittsburgh are especially significant. Prior to Weed and Seed, the Crawford-Roberts area had a drug arrest rate four times higher than the rest of the city, but 4 years after Weed and Seed began, the target area's drug arrest rate was about the same as the rate for the rest of the city. Rates for Part 1 arrests also declined in Hartford, Pittsburgh, North Manatee, South Manatee, and Seattle. Crime trends in the target areas A reduction in crime is generally viewed as one of the most important goals of Weed and Seed. Thus, the extent to which crime was reduced in the target areas is an important indicator of Weed and Seed's overall success, recognizing, of course, that a variety of factors unrelated to Weed and Seed can affect crime rates. Exhibit 3.3 shows annual Part 1 crime statistics for the Weed and Seed target areas and for the remaining non-Weed and Seed part of the jurisdiction. The format of this exhibit is the same as the Part 1 arrest and drug arrest charts shown in exhibits 3.1 and 3.2. As was the case with arrest data, Salt Lake City is presented differently because only 5 months of crime data were available prior to the start of its Weed and Seed program in August 1995. The annual number of Part 1 crimes in the target areas shown in exhibit 3.3 highlights the significant differences across the target areas--there is roughly a 10-fold difference in the number of Part 1 crimes between the target areas in the low range (Hartford and North Manatee) and those in the high range (Seattle and Shreveport). The Part 1 crime rates per 1,000 residents vary less, with about a threefold difference between the smallest (Akron and North Manatee) and largest (Meadows Village and Shreveport) rates. A primary area of interest in exhibit 3.3 is the difference between the target area crime rates before and after the start of Weed and Seed. Exhibits 3.4 and 3.6 highlight these differences, with exhibit 3.4 comparing the Part 1 crime rate per 1,000 residents in the year before Weed and Seed to the first year of Weed and Seed and exhibit 3.6 comparing the Part 1 crime rate in the year before Weed and Seed to the second year of Weed and Seed.[7] Before discussing the numbers in these exhibits, three comments should be made. o In each exhibit, the sites are ordered from the largest percentage decrease in Part 1 crime on the left to the largest percentage increase on the right. o Akron is the only site for which researchers did not have 2 complete years of crime data following the start of Weed and Seed; thus, this site is not included in exhibit 3.6. o Salt Lake City's percentage changes are based on 5-month, rather than 12-month, periods. Exhibits 3.4 and 3.6 show Part 1 crime decreases in most of the Weed and Seed target areas. Comparing the first year of Weed and Seed to the year prior to its beginning (exhibit 3.4), 7 of 10 target areas experienced decreases (the exceptions being Seattle, South Manatee, and Salt Lake City), and 7 of 10 target areas also experienced fewer Part 1 crimes in the second year of Weed and Seed (exhibit 3.6) than in the year prior to its beginning. The areas with increases over this period were South Manatee, Meadows Village, and Salt Lake City. Eight of nine target areas (all areas except Meadows Village) showed a decrease in Part 1 crime from the first year of Weed and Seed to the second year. In both sets of comparisons, the Stowe Village target area in Hartford experienced the largest percentage decrease in Part 1 crime across all the target areas. Perhaps the most striking feature of exhibits 3.4 and 3.6 is the wide range of Part 1 crime changes across the entire set of Weed and Seed target areas, particularly after 2 years of implementation. The percentage changes over 2 years range from a 46-percent decrease in Hartford's Stowe Village to a 14-percent increase in Salt Lake City. Four other target areas had double-digit percentage decreases (Crawford-Roberts in Pittsburgh, 24 percent; North Manatee, 18 percent; Shreveport, 11 percent; and the Central District in Seattle, 10 percent). One site (West Las Vegas) had a single-digit decrease (6 percent), and three target areas experienced increases in Part 1 crime (South Manatee, 2 percent; Meadows Village, 9 percent; and Salt Lake City, 14 percent). Section 3.4 in this report attempts to place these widely varying changes in some perspective by considering a range of possible contributing factors. A final comment on exhibits 3.4 and 3.6 should be made regarding the Salt Lake City target area, which had the largest crime increase in both exhibits. As noted above, only 5 months of crime data prior to Weed and Seed were available from this site; thus, year-to-year comparisons were based only on the March-to-July time period. The year-to-year percentage changes in Part 1 crime may in fact have been very different if information from the entire 12-month pre-Weed and Seed period had been available. Moreover, the lowest monthly Part 1 crime total over the entire period for which there are data was April 1995 (the middle of the city's pre-Weed and Seed period). Weed and Seed began amid an upsurge in crime that started after that month and continued through October 1995. As an additional indicator of crime-related outcomes over time, crime trends beyond the first 2 years of Weed and Seed implementation can be seen in the Pittsburgh and Seattle numbers in exhibit 3.3, both of which began Weed and Seed programs at least 2 years prior to the other sites in the national evaluation. In Pittsburgh, the Part 1 crime rate in the fifth year following the start of Weed and Seed in Crawford-Roberts was 44 percent lower than the year prior to Weed and Seed (compared to 24 percent lower after 2 years). In the Central District target area in Seattle, the Part 1 crime rate in the fourth year following the start of Weed and Seed was 22 percent lower than the year prior to Weed and Seed (compared to 10 percent lower after 2 years). Possible contributing factors Criminologists, law enforcement personnel, and city officials have offered a variety of explanations for the nationwide drop in crime, including changing demographics, community policing, a healthy economy, and increasing incarceration. Explaining crime rate changes in the Weed and Seed target areas must obviously consider two additional factors--the Weed and Seed program and other law enforcement or social programs that either existed in the target areas prior to Weed and Seed or that started during Weed and Seed. That Part 1 crime dropped in the target areas cannot be disputed; however, it is not possible to state definitively the extent to which different factors contributed to the decrease. That said, it is nevertheless useful to qualitatively discuss possible contributing factors. As discussed in this section, a number of factors have a strong correlation with the observed changes in crime. Crime trends in nontarget areas By design, the Weed and Seed target areas constitute a small part of the jurisdictions to which they belong. Three of the target areas contain less than 1 percent of the jurisdiction's population; Salt Lake City's target area contains 12 percent of the city's population, the highest percentage among all the target areas. Given the size of the target area relative to the entire jurisdiction, it is reasonable to assume that crime trends in the nontarget areas of the jurisdiction would have some effect on the trends in the target areas. Part 1 crime rates in the target areas can be compared to the Part 1 crime rates in the nontarget parts of the jurisdiction by examining exhibits 3.4 through 3.7. Exhibits 3.4 and 3.6 show crime trend figures for the target areas, while exhibits 3.5 and 3.7 show analogous figures for the nontarget areas. Exhibits 3.4 (trends in the target areas after 1 year of implementation) and 3.5 (trends in the nontarget areas after 1 year of implementation) show that, after 1 year of Weed and Seed implementation, in 7 of 10 target areas--Akron, Hartford, Pittsburgh (Crawford-Roberts), Las Vegas (both Meadows Village and West Las Vegas), North Manatee, and Shreveport-- the Part 1 crime rate improved more in the target area than in the rest of the corresponding jurisdiction. Exhibits 3.6 (trends in the target areas after 2 years of implementation) and 3.7 (trends in the nontarget areas after 2 years of implementation) show that, after 2 years of Weed and Seed implementation, in 6 of 9 target areas--Hartford, Pittsburgh (Crawford-Roberts), North Manatee, South Manatee, Shreveport, and West Las Vegas--the Part 1 crime rates improved more in the target area than in the rest of the corresponding jurisdiction. The Weed and Seed target areas, because they are so much smaller geographically than their corresponding nontarget areas, show greater year-to-year variation in crime rates. Still, one should note the magnitude of the difference between the Part 1 crime rate change in Stowe Village in Hartford and in the rest of Hartford, compared with the North Manatee target area and the rest of Manatee County. After 2 years of implementation, the Stowe Village Part 1 crime rate had dropped 46 percent, compared to 21 percent for the rest of the city (a 25-point differential). The North Manatee Part 1 crime rate dropped 18 percent, compared to an 8-percent increase in the nontarget section of the county (a 26-point differential). Also noteworthy is the pattern of target area changes in those sites whose nontarget areas experienced large decreases or increases in crime. The two sites with the largest Part 1 crime decreases in the nontarget areas (Hartford and Pittsburgh) are the same two sites where target areas achieved the largest Part 1 decreases. At the same time, the site with the largest Part 1 crime increase in its target area--Salt Lake City--also exhibited the largest Part 1 crime increase of all nontarget areas. Preexisting crime trends in the target areas Given that most of the Weed and Seed target areas experienced declining crime rates during the implementation period, it is important to ask whether these trends represent the continuation of a preexisting downward crime trend or whether the Weed and Seed period coincided with the reversal of an upward trend. The Part 1 crime statistics shown in exhibit 3.3 only partially answer this question, since only 1 complete year of pre-Weed and Seed crime data are shown for Las Vegas, Manatee, and Pittsburgh. For target areas in these sites, and to some extent for other target areas as well, a review of the monthly Part 1 crime trends in the case study documents creates a more complete picture of preimplementation and postimplementation trends. The general preimplementation and postimplementation Part 1 crime trends cast Weed and Seed in an especially favorable light in two sites-- Pittsburgh and Shreveport. At both sites, the implementation of Weed and Seed coincided with a marked reversal of an upward crime trend in the target areas. In one other site--Stowe Village in Hartford--there was a preexisting downward trend in Part 1 crime, but the onset of Weed and Seed coincided with an acceleration of the downward trend. In the Akron target area, West Las Vegas, North Manatee, South Manatee, and the Seattle target area, the preexisting downward trend continued at roughly the same rate over the Weed and Seed implementation period. Finally, both Meadows Village in Las Vegas and the Salt Lake City target area experienced overall increases both before and after the start of Weed and Seed. Program intensity Program intensity refers to the concentration of program resources. A possible measure of program intensity is the level of resources per capita or per unit area. However, the total level of program resources directed to the target areas, including Federal Weed and Seed funds, other Federal funds, State and local funds, and community-based resources, is not known. The only monetary figures available are the Federal Weed and Seed awards, which are shown in exhibit 2.1. This exhibit shows the gross levels of Federal Weed and Seed funding at the eight sites, which can be divided into two groups according to their funding status at the start of the national evaluation--the six demonstration sites and the two officially recognized sites (Akron and Salt Lake City). In addition to not accounting for nontarget funds, the figures in exhibit 2.1 are site-specific and not target area-specific. Thus, while crime data are target area-specific, Federal Weed and Seed spending figures are not available by target areas in Las Vegas, Manatee/Sarasota, and Pittsburgh, or for Seattle, which expanded Weed and Seed beyond the original target area boundaries. The population and area of each target area are available, however, and are shown in exhibit 1.2. As shown, the target area populations range from 1,300 in Hartford to 24,000 in Akron; target area sizes range from 0.1 square mile in Hartford to 6.3 square miles in Salt Lake City. As was the case in the discussion of the relationship between crime trends in the target area and the rest of the corresponding jurisdictions, some relationship appears to exist between crime trends and program intensity at those sites with the largest crime increases or decreases. For those sites between these extremes, the pattern is less clear. Hartford, for example, experienced the largest Part 1 crime decrease among all the target areas. At the same time, it is also the smallest in terms of both population and area. Moreover, unlike Crawford-Roberts in Pittsburgh and North Manatee, which only received part of the site's Federal Weed and Seed award, Stowe Village remained Hartford's sole target area throughout the program. Salt Lake City, on the other hand, which experienced the largest increase in Part 1 crimes, also has the largest target area in square miles and, along with Akron, the smallest level of Federal Weed and Seed funding. Arrest volume A final possible contributing factor is the volume of arrests made in the target areas. In the case of interventions that are primarily arrest-oriented, statistical tests could be performed to determine whether the arrest volume in one time period leads to a subsequent drop in crime. Given the multidimensional nature of Weed and Seed interventions, statistical models attempting to link program effort to program impact would grossly simplify reality. The approach taken here is to examine, in a qualitative way, the relationship between arrests and Part 1 crime. Perhaps the most distinguishing feature of the target area arrest statistics shown in exhibits 3.1 and 3.2 is that Hartford's arrest rates, especially drug arrests, are so much higher than in other target areas. Thus, Hartford had the highest Part 1 and drug arrest rates and also the largest decrease in Part 1 crime of all target areas. At the same time, as noted earlier, Hartford's target area has the smallest population and area, and Hartford experienced the largest jurisdictionwide drop in crime. Although there are no clear trends between Part 1 arrests and Part 1 crimes across the sites, changes in the drug arrest rates appear to be associated with changes in the Part 1 crime rate. For example, among these six target areas for which researchers have arrest data, the four with decreases in Part 1 crime (Hartford, Pittsburgh, North Manatee, and Shreveport) all experienced initial high rates of drug arrests, suggesting an initial period of intense weeding activities, followed by declining drug arrest rates. Assuming the level of enforcement as measured by police presence has remained somewhat constant, this trend reflects success in reducing drug activity. Salt Lake City and South Manatee, for which researchers have arrest data that show increases in Part 1 crime, have not shown a decrease in their drug arrest rates. Indeed, both of these target areas had large increases in drug arrests in the last reporting period shown in exhibit 2.2, suggesting, perhaps, that these sites have not yet succeeded in reducing the level of drug activity in the target areas. Thus, across these six sites, the changes in drug arrest volumes follow the same general pattern as the changes in Part 1 crime. Notes 1. Geocoding is a process that involves assigning geographic coordinates (latitude and longitude) to an address, in this case either the address where the crime occurred or where the arrest was made. Once records are geocoded, mapping software can be used to extract those arrests and crimes that occurred within the Weed and Seed target area. 2. The definition of a drug arrest used in these analyses is an arrest for which at least one charge explicitly involves drugs (for example, possession with intent to sell narcotics). 3. An alternative outcome measure, which focuses more on the quality, as opposed to the quantity, of arrests made, is the period of incarceration to which arrestees were sentenced. As noted in chapter 2, prosecution and court data were generally not available. 4. The source of the incident-level arrest data, on which the analyses in this section are based, was the local law enforcement agency; thus, the figures do not include arrests made by Federal law enforcement agencies that were processed through Federal arrest booking facilities. 5. Data for partial years both before and after the time period indicated in exhibit 3.1 are included in the arrest trends in the individual case studies. 6. In 1997, a new law enforcement records management system was installed in Salt Lake City. Only the automated records after March 1995 were converted to the new system. 7. Exhibits 3.5 and 3.7 contain Part 1 crime rates for the non-Weed and Seed parts of the jurisdiction. The ordering and placement of exhibits 3.4 through 3.7 in this document are designed to facilitate comparisons between the target areas and the non-Weed and Seed parts of the jurisdiction. ------------------------------- Chapter 4 An independent agency surveyed residents in the Weed and Seed target areas in 1995 and again in 1997 to measure their perceptions of crime, public safety, police performance, general quality of life, and awareness of Weed and Seed itself. The opinions of nonelderly, long-term residents were compared over the 2-year period to determine what, if any, effect Weed and Seed had on their satisfaction with the community. Evaluators also interviewed participants in four types of seeding programs, including youth recreation, community building, employment and training, and violence prevention. According to participants, these programs provided services that would otherwise not have been available and increased feelings of overall security. Perceptions of Community Residents and Program Participants In all eight sites that participated in this evaluation, a survey of target area residents was conducted at two separate time intervals--from March to July 1995 (by the Institute for Social Analysis) and from December 1997 to January 1998 (by Abt Associates Inc.). The following discussion refers to these data collection efforts as the 1995 and 1997 surveys, respectively. The objective of the survey data collection and analysis was to measure changes in residents' perceptions of public safety, crime, police performance, general neighborhood quality of life, and awareness of the Weed and Seed program. In the interest of comparing the findings obtained from the two surveys, the 1997 survey was designed with the following features: o For each site, the geographical boundaries of the survey area were the same as in 1995.[1] The survey areas by site were as follows: Akron, West Side; Hartford, Stowe Village; Las Vegas, West Las Vegas; Manatee/Sarasota, North Manatee; Pittsburgh, Hill District (Crawford-Roberts neighborhood); Salt Lake City, West Side; Seattle, Central District; and Shreveport, Highland/Stoner Hill. o The wording of questions from the 1995 survey was retained verbatim in 1997. (Refer to the questionnaires shown in appendixes 1 and 2, respectively.) For selected questions, additional response categories were added in 1997 to provide a more complete range of possible responses. For these items, care was taken during analysis to aggregate responses in ways that would preserve the comparability of the findings between 1995 and 1997.[2] o There were also some notable differences in the methods used in the two surveys. The 1995 survey consisted of inperson interviews, based on city-provided address lists. The 1997 interviews were conducted by telephone, based on listed telephone numbers for residential addresses within the survey area. The 1995 survey consisted of 83 substantive items. The 1997 survey included only a subset (31) of these 83 substantive items. (For both surveys, the count excludes items related to respondent demographic characteristics and other basic interview data.) The 1995 interviews required 30 to 40 minutes. The 1997 interviews typically lasted 12 to 15 minutes. The decision to proceed in 1997 with telephone interviews and a shortened instrument was based on the difficulties experienced in 1995 in completing the targeted number of 400 interviews per site. (In none of the sites was this target reached.) The 1997 survey design called for 300 completed interviews per site. Distribution of respondents by age and length of neighborhood residence The methodological differences between the two surveys, combined with underlying population changes in the evaluation sites, led to a shift between 1995 and 1997 in the demographic profile of respondents in each surveyed target area. Exhibit 4.1 shows the distribution of respondents in each site by two key characteristics: age and length of residence. Those 60 years of age or older at the time of the survey were defined as elderly; those less than 60, nonelderly. Those residing in the target area for more than 2 years were defined as long-term residents; those for less than 2 years, short-term residents. Two options were considered to account for the demographic shift in the surveyed population and make the precomparison/postcomparison as valid as possible. One option was a multivariate regression approach for modeling the responses to each survey question as a function of the respondent's demographic characteristics and an indicator of the survey year. This was rejected on the basis of insufficient numbers of observations in each site to reliably estimate such a model. A second option, which was ultimately selected, was to focus on a demographically comparable subset of respondents from both surveys: the nonelderly, long-term residents. This group constituted 50 percent of total respondents in 1995 and 49 percent in 1997. Because respondent age typically is a significant factor in responses to survey questions of this type, a shifting age profile over time could cause a change in the pattern of responses at any site. Elderly residents as a share of total respondents across all eight sites rose from 29 percent in 1995 to 35 percent in 1997. For some sites--Akron and Pittsburgh, in particular-- this shift in age distribution was even more pronounced. It became clear that any survey analysis that did not account for respondents' ages would likely confound the effects of demographic change with the effects of Weed and Seed in the target area. At the same time, any separate analysis of community perceptions among elderly residents was hindered by the small numbers of elderly respondents in some sites--even when using 60 years (rather than 65 years) as the threshold age. Hartford had fewer than 10 elderly respondents in both 1995 and 1997; three other sites (Akron, Salt Lake City, and Shreveport) had 30 or fewer elderly respondents in 1995. In focusing this analysis on nonelderly residents, it was then decided to exclude short-term residents for several reasons. First, some questions of primary interest to the evaluation asked the respondent to assess community conditions compared to 2 years ago or to report victimization over the past 2 years. Second, even on other questions, the length of residence was found to be a significant factor in the pattern of responses and thus could confound the interpretation of findings with respect to Weed and Seed. Any separate analysis of the short-term nonelderly residents was precluded by their small numbers--fewer than 30 respondents in both 1995 and 1997 for three sites (Akron, Hartford, and Manatee/Sarasota), with two other sites (Pittsburgh and Salt Lake) having fewer than 30 such respondents in 1995. All of the findings that follow thus pertain to the survey responses of nonelderly long-term residents. For all sites, there were more than 30 such respondents in both 1995 and 1997. In three sites (Las Vegas, Pittsburgh, and Seattle), there were more than 100 such respondents in both years. To examine the possible impact of Weed and Seed on neighborhood residents' perceptions, the responses of nonelderly long-term residents between 1995 and 1997 were compared question by question. Significant differences between the two surveys are shown by site and by question in appendix 3.[3] Personal safety and neighborhood satisfaction The first survey questions displayed in appendix 3 pertain to perceptions of personal safety and general satisfaction with the neighborhood. Researchers focused on changes in the proportion giving the most positive response, reporting they feel either "very safe" or "very satisfied." Pittsburgh was the only site to show a significant increase between 1995 and 1997 in the percentage of nonelderly long-term residents indicating they feel very safe out alone in their neighborhood during the day. (No site showed a significant change between 1995 and 1997 on the survey question regarding personal safety out alone in their neighborhood after dark.) Similarly, Pittsburgh was the only site to show a significant increase in the percentage of nonelderly long-term residents who were "very satisfied" with their neighborhood as a place to live. Severity of neighborhood crime. Survey respondents were also asked if they perceived specific crimes in their neighborhood to be a "big problem" versus a lesser problem. For crimes related to drug sales or drug use, depending on the particular type of illegal activity, four of the eight sites showed some significant reduction between 1995 and 1997 in the percentage of respondents indicating that such crimes were a big problem. These four sites (Akron, Manatee/Sarasota, Pittsburgh, and Seattle) all showed reductions between 1995 and 1997 in the perceived severity of "drug dealers on streets, street corners, or in other public places." Three of these sites (Akron, Manatee/Sarasota, and Seattle) also had significantly fewer respondents in 1997 than in 1995 who indicated "drug sales out of homes or apartments" were a big problem in their neighborhood. Respondents in three of these sites (Manatee/Sarasota, Pittsburgh, and Seattle) were also significantly less likely in 1997 than in 1995 to say "drug sales" were a big problem in their neighborhood. With regard to other types of crimes, Manatee/Sarasota was the only site that showed any significant reduction between 1995 and 1997 in the perceived severity of either "burglaries and other property crime" or "robbery and other street crime." For "violent crimes (such as shootings, assault, and so forth)," there was a significant reduction in perceived severity in Hartford, Manatee/Sarasota, and Pittsburgh between 1997 and 1995. Similarly, the perception of "gang activity" as a big problem was significantly lower in 1997 than in 1995 for both Hartford and Pittsburgh. Victimization in the neighborhood. In addition to questions regarding the perceived severity of crime in their neighborhoods, residents were asked whether they (or a member of their family) had been a victim of crime over the past 2 years. For instance, residents were asked whether anyone had "broken into their home, garage, or another building on your property to steal something." Only in Manatee/Sarasota were nonelderly long-term residents significantly less likely in 1997 than in 1995 to indicate such victimization. This site, along with both Las Vegas and Pittsburgh, also showed a significantly lower percentage of respondents who had been "beaten up, attacked, or hit with something such as a rock or bottle." Pittsburgh was also the only site in which nonelderly long-term respondents were less likely in 1997 than in 1995 to have been "knifed, shot at, or attacked with some other weapon" within the past 2 years. In none of the sites was there a significant change between 1995 and 1997 in the percentage of respondents who had been victims of robbery over the last 2 years. Police effectiveness and responsiveness. Of the four sites that showed significant reductions in the perceived severity of drug-related crime, Manatee/Sarasota, Pittsburgh, and Salt Lake City showed a significant increase in the percentage of respondents who said the police were doing "a very good job of keeping order on the streets and sidewalks in this neighborhood." Pittsburgh respondents were also more likely in 1997 than in 1995 to indicate police were doing a very good job at "controlling the street sale and use of illegal drugs." As evidence of increased police presence in the neighborhood, the nonelderly long-term respondents in Salt Lake City and Seattle were significantly more likely in 1997 than in 1995 to "have seen a police officer chatting or having a friendly conversation with people in the neighborhood." For Seattle respondents, however, this appears to contradict the finding that they were less likely to perceive the police as "very responsive" to community concerns. Only in Hartford was there a significant increase in the percentage of respondents indicating the police were "very responsive" to community concerns in the neighborhood. Awareness of Weed and Seed. In all sites except Akron, there was a significant increase between 1995 and 1997 in the percentage of nonelderly long-term respondents who indicated that they "have heard of a program called the Weed and Seed Program." The lack of program awareness in Akron is understandable, given its much smaller scale and more recent implementation. General neighborhood quality of life. A meaningful indication of the changed living conditions in the Weed and Seed target areas is the pattern of responses to the following survey question: "In the past 2 years, would you say this neighborhood has become a better place to live, a worse place to live, or stayed about the same?" Exhibit 4.2 shows the percentage distribution of responses by site, for nonelderly long-term respondents. The 1995 responses correspond to perceived change in neighborhood conditions during 1994-95; the 1997 responses indicate perceived changes during 1996-97. It is useful to interpret these responses in light of Weed and Seed's start date for each site, which is also shown in the exhibit. One can divide the sites into three different groups as follows: o Hartford and Seattle--where more than 25 percent of respondents indicated the neighborhood had become a better place to live in both the 1995 and 1997 surveys. (In fact, in these two sites the corresponding percentages approached or exceeded 50 percent in both surveys.) o Las Vegas, Pittsburgh, and Shreveport--where the percentage of respondents indicating the neighborhood had become a better place to live was 25 percent or less in the 1995 survey but more than 25 percent in the 1997 survey. (In each of these sites, the difference in the percentages between 1995 and 1997 was statistically significant.) o Akron, Manatee/Sarasota, and Salt Lake City--where less than 25 percent of respondents indicated that the neighborhood had become a better place to live in both the 1995 and 1997 surveys. The placement of Akron and Salt Lake City in the last group might be explained by the relatively recent implementation of Weed and Seed in these target areas. For Manatee/Sarasota, where the survey findings might have suggested general improvement in the perceived quality of life, the distressed local economy may be responsible. Conversely, the placement of Seattle in the first group, and Las Vegas and Shreveport in the second group, may reflect the improving economic conditions in these metropolitan areas, as earlier noted survey evidence did not indicate any significant favorable shifts in perceived personal safety or police responsiveness. For the remaining two sites--Hartford and Pittsburgh--the relatively high and rising perceptions of neighborhood quality of life appear consistent with the pattern of responses to other survey questions, which indicated a lessening in the perceived severity of drug-related and violent crimes. General assessment. What overall interpretation should one give these survey findings? Chapter 5 addresses this issue at some length, in the context of the implementation experience and crime trends in target areas. At this point, however, several observations deserve mention: o Where significant changes in the perceptions of residents were observed between 1995 and 1997, the changes were in the favorable direction, almost without exception. o The choice of a threshold for statistical significance is ultimately arbitrary. Some may consider the research team's use of the conventional standard of 10-percent significance or better too strict. (The detailed tabulation displayed in appendix 3 allows the reader to apply alternative standards.) o The lack of significance on some questions, especially in sites (such as Hartford) where one might expect the target area's reduced crime rate to influence residents' perceptions, may reflect a variety of mitigating factors. First, it may be too early for residents to perceive the improved crime situation, especially where the program was more recently implemented (as in Akron). Second, where the program was implemented early (as in Seattle), significant changes in residents' perceptions may have already occurred by the time of the 1995 survey. Third, in some target areas, the 1997 responses may reflect a reversal of earlier improvements in the crime rate (as occurred in late 1996 or early 1997 in Akron, Hartford, and Seattle). o The longest term community residents, those who have resided in the target area for many years, may have difficulty perceiving recent changes, as their feelings about the neighborhood may be colored by unfavorable experiences, which long preceded Weed and Seed. With these caveats in mind, the sites appear to cluster into three groups with respect to the overall pattern of survey findings for nonelderly, long-term residents, comparing 1995 and 1997: o Manatee/Sarasota (North Manatee) and Pittsburgh (Crawford-Roberts) exhibited substantial evidence of changes in residents' perceptions across multiple outcome measures, including the severity of both drug-related and violent crime and police effectiveness in controlling street crime. o Akron, Hartford, and Seattle exhibited some evidence of changes in residents' perceptions on selected crime dimensions, either drug-related crime (Akron and Seattle) or violent and gang-related crime (Hartford). o Las Vegas (West Las Vegas), Salt Lake City, and Shreveport exhibited little evidence of changes in residents' perceptions of general public safety or the severity of specific types of crime in the neighborhood. Participant interviews To gain the perspective of individuals in the community whom the seeding programs were intended to benefit, interviews were conducted with 40 to 60 program participants in Pittsburgh, Salt Lake City, Seattle, Hartford, Manatee, Shreveport, and Las Vegas. No interviews were conducted in Akron because there were no seeding programs at the time. Evaluation staff visited 4 to 6 programs in each of the sites and conducted interviews in groups of 7 to 10 people. These interviews were not intended to measure outcomes of the programs that were visited; rather, they were designed to learn the perceived benefits and drawbacks of the programs from individuals who participated in them. This analysis is about perceptions and not impacts, but there is no way to be certain the benefits or hindrances these people experienced while in the programs are explicitly due to their participation. Further, the analysis is meant to present individual perceptions based on time-limited involvement in these specific programs; long-term effects of these programs on the lives of the program participants cannot be deduced from the interviews conducted. It is also important to note the interviews were arranged by each site's Weed and Seed coordinators, who worked with the program directors to arrange the time, place, and individuals to be interviewed. Interviewees were selected at the discretion of the program managers, based on participants' availability and their willingness to participate in the interview process. (The interviews should not be interpreted as representative of all participants.) The seeding programs that were visited fall under four general categories: youth recreation, community building, employment and training, and violence prevention. The following is a synopsis of cross-site themes that emerged from the participant interviews for each program category, including descriptive quotations that illustrate participant perspectives. Youth recreation and education programs The purpose of these programs is to give young people opportunities to participate in activities after school when many youths typically would be without adult supervision. The programs visited included the Police Activities (or Athletic) League, Boys and Girls Club Athletic Programs, Safe Haven AfterSchool Programs, and Youth Enrichment Programs. Depending on the age group served, these programs can provide a safe haven for children who are afraid to be on the street, a place where they can receive help with their homework, and/or a place where they can participate in recreational athletic activities after school, instead of sitting at home and watching television. This was the most prevalent category of programs visited. Many individuals interviewed noted the lack of quality programming in their communities prior to Weed and Seed, commenting that without these recreation programs they would have nothing to do after school. Youth recreation programs offer young people ways to fill a void in the day that may otherwise provide temptations or risks. A number of the participants in various communities told evaluators if it were not for these types of programs, they would be "getting into trouble," "doing drugs," or just "hanging around their house and watching television." Further, many of the participants indicated that their program leaders were role models for themselves and other neighborhood youths. In addition, virtually all individuals said participation in recreational activity has improved their school performance, including helping them achieve higher grades. Those interviewed generally felt that homework assistance, provided in many of these programs, was the aspect of the youth recreation programs that most helped their performance in school. One child in Shreveport said, "They make sure you do your homework; if you don't, you can get in trouble. At home, my mama, she don't never help me." Others noted the additional structure the program adds to their lives. Overall, the children seemed pleased with the discipline and camaraderie many of these programs provide. Parents of children who participated in afterschool recreation programs reported the primary benefit to be improved school performance. Further, they felt the programs helped their children to be versatile by involving them in various activities which kept them constructively engaged. One Seattle parent told evaluators, "It helps when the kids get out of school and gives them somewhere to go and do their homework and help them with activities. While if I get home from work at 6 o'clock, it is time to fix dinner, get them ready for school the next day and that is it--there isn't enough time for homework." Additionally, another mother in Seattle had this to say about her daughter's recreation program: "My daughter loves coming here, and I would rather have her here where she likes it and feels safe than being at home alone all afternoon. I think this place has made a difference with the children because most of them would be latchkey children. With this program there are not so many children at home. Kids have a more positive outlook on things as opposed to just hanging out and not doing anything productive. It gives them someplace to go where there is structure and respect." Although this particular woman's medical story is unique, the benefits she reported were common among nearly all interviewed. This mother had two daughters, ages 10 and 12, in the Police Activities League (PAL) Summer Program, which is a recreation program for Manatee youths. The mother had suffered a stroke and was concerned that her children would be idle while she recuperated. She wanted her children to be engaged in structured, productive activities and saw a lack of supervision in other programs. At PAL, teachers and police officers provide instruction. She said most of the police officers are male, and the kids stay on their toes because of their authority and uniforms. She said her children have grown since participating in PAL. Before, her daughters were harder to manage, fussy, and bored. Her eldest daughter only wanted to hang out with her peers and was starting to go "the wrong way." Now, her daughters have "totally changed" their attitudes and have better morale. She said her oldest daughter, who is an artist, participated in the summer art program and received an outside art scholarship with staff assistance. She also became an A student and joined the PAL Academy instead of public school. The youngest daughter is now on the PAL basketball team and loves sports. Prior to her participation, she did not want to do anything, according to her mother. This mother knows the police officers who work with her daughters at PAL and has received valuable advice from staff members. She said, "I don't think I could have made it through without having their support." Community-building programs Community-building programs are designed to enable residents to take charge of their own safety and the safety of their communities. Examples of programs visited under this category include neighborhood and crime watch programs such as Mobile Watch. These programs involve volunteer community residents patrolling their neighborhoods in marked vehicles to look for and report suspicious or criminal activity. Often, these programs also provide a mechanism for community members to work with and get to know the police officers assigned to their neighborhoods. Participants in the community-building programs tended to be involved and interested in affecting some change in their neighborhoods. Many reported that their neighborhood or crime watch programs provided them with a feeling of control over their own safety. Before participating in these programs, many interviewees reported feeling unsettled and afraid to go out alone in their own neighborhoods. Participants also indicated that they enjoyed positive relationships with many of the police officers on their beat, and they believed the activities of the community groups helped officers make arrests. One man who participated in a community-building program in Salt Lake City considered his Mobile Watch program a bridge for improving relations between citizens and police officers. Program members generally felt residents were more comfortable sharing information about criminal activity with the neighborhood watch rather than with flagging down a police officer, who often was too busy to deal with everyday individual concerns. Police in many target areas often must focus on violent and dangerous crimes, leaving them little time for nonviolent problems that annoy community residents, such as noise disturbances. Members of community-building groups, on the other hand, can spend the time collecting data, observing, and documenting events. Participants in this community believed this effort resulted in greater police attention to the community's concerns. All participants reported that they perceived a noticeable reduction of drug and criminal activity in their patrol areas. They felt safer in their communities--despite other statistical indications--and that is "all that matters to them." Further, many individuals across sites indicated that they now know their neighbors, making it easier to notice when someone suspicious is in their neighborhood. One woman who had children said her participation in this type of program made her feel that if something happened, her children could now go to one of the neighbors for help. Finally, participants also felt there was high awareness of their particular programs in their neighborhoods. Employment and training programs The purpose of employment and training programs is to help train community residents and place them in jobs. The programs work with a range of people, from high school students to adults who want to change their career focus. Most programs offer a range of services, including Graduate Equivalency Degree training, computer classes, English as a second language (ESL), and medical assistant classes. Once participants have completed their coursework, most agencies have a career center that helps place them in jobs. One of the most obvious benefits of employment and training programs is the income the individuals receive when they start working. Many interviewees said they previously had been on transitional assistance and were relieved to no longer have to "deal with the welfare office." Further, the increased self-esteem that comes with finding successful employment was brought up on a number of occasions. One interviewee from Seattle exclaimed, "I never thought I could know so much. I am impressed with myself; the teachers here are great and very patient." Another Seattle participant said he felt he was a role model for his peers. After seeing what he accomplished, many of his friends enrolled in the same vocational program. Further, he explained, "When people have jobs, and therefore something income generating and productive to fill their days, they are less likely to get involved in criminal activity." Additionally, one of his classmates said, "There are lots of people who would be jail-bound before this program, and some of my friends are so impressed with me that they are starting to come here, also." Parents of children in a summer work program in Manatee thought their children had very positive work experiences and had learned responsibility, including budgeting the money they earned. One woman's son bought his first pair of sneakers with his earnings. Another parent explained that the experience was educational for her children, teaching them responsibility and building their skills and confidence. Violence prevention programs Violence prevention programs work with young people and adults to help end violence that is often present in their communities. Examples of programs include school mediation programs, peer advisory programs, and youth and family specialist programs, which work with participants to peacefully resolve conflicts and deal with anger and frustration. Participants often later serve as role models in the community and teach others what they have learned. Particularly for teens, these programs help those who have trouble dealing with anger. One teen in Salt Lake City explained that since working with a youth and family specialist, she had gotten into less trouble. When she was about to do something harmful, she called her specialist. Her specialist then talked her through the problem and helped her find alternative ways to deal with her feelings. She said her specialist was always there for her, night or day, and "bailed her out" on a number of occasions. Another participant from Salt Lake City said his youth and family specialist was extremely helpful. He appreciated that the specialist visited him once a week to make sure he was not getting into trouble. The specialist also provided the participant with his beeper number. If the young man felt he was headed toward trouble, he always beeped his specialist before acting on his anger. Finally, many participants told evaluators they had changed their future plans because of these programs. For example, one girl who was promised a job by her program director when she "cleaned herself up" explained that she looked forward to getting a job and not just "kickin' it" with her friends. Participant suggestions for program improvement Following is a selection of suggestions provided by the participants in various programs in each of the sites: One Manatee County participant thought her program should be advertised more; she would not have known about it if she had not known the program director. Participants said that, in the black community, communication through churches would be a good way to inform people about summer jobs and tutoring. Another Salt Lake City woman said there were not enough activities for children outside of school. She knew of no other organized programs or activities. The children would play outside on the playground and hang out with older kids, who could be a bad influence.[4] Suggestions for improvements in the community-building programs often focused on training. Many of those interviewed felt more extensive training was needed before people became participants in these programs. Many believed additional training would promote the safety of volunteers. In addition, many felt that, although positive relationships existed with police, increased communication with police officers would help their cause. Many participants found when they communicated with police, their concerns often were more effectively addressed. Summary overview Based on participant comments, seeding programs appear to provide services that would otherwise not be available in the target areas. Most of those interviewed indicated that participation in the seeding programs helped them feel more secure emotionally, physically, or both. The general themes that emerged revolved around providing additional structure and discipline for target area youths and providing opportunities and assistance for adults to work toward personal and professional growth. Notes 1. In Akron, Hartford, Salt Lake City, Seattle, and Shreveport, the survey area consisted of the entire Weed and Seed target area under study in this evaluation. Because the 1995 survey was conducted in only one portion of the target area for Las Vegas, Pittsburgh, and Manatee/Sarasota, it was decided to retain the same designated areas in 1997 to ensure consistent comparisons. 2. For example, in questions on "how good a job are the police doing" in different aspects of law enforcement, the 1995 survey allowed the respondent to indicate "a very good job, a good job, a fair job, or a poor job." The 1997 survey also allowed the respondent to indicate "a very poor job." The findings aggregated the "poor job" and "very poor job" responses for 1997 before comparing the pattern of responses with 1995 results. 3. For each site, the difference between 1995 and 1997 in the response pattern to any given question was regarded as significant only if it met the conventional statistical standard of 10-percent significance, based on a t-test for differences in proportions. 4. It is important to note that in some Weed and Seed communities, participants stated that over the past few years, youth recreation programs have increased dramatically, to a level with which they are satisfied. ------------------------------ Chapter 5 Although it remains impossible to precisely state what the effect of Weed and Seed has been given the number of programs and other funding already at work in the target areas, the overall impact has been positive. However, it was clear that several factors, including the strength of the existing social and institutional infrastructure, severity of crime, economic development advantages, and transiency, influenced success. Program design, implementation procedures, concentration of resources, and leadership also proved essential to a positive outcome. Comparisons of target areas among the various sites and with nontarget areas provided suitable backdrops for assessment. Assessing the Weed and Seed Strategy The most important questions to be answered are whether the sites achieved the intended changes in measurable crime-related outcomes, and what factors appear to promote successful implementation and achievement of intended results. As noted in chapter 1 and explained further in this section, it is extremely difficult to draw definitive conclusions about the impact of Weed and Seed. However, it was possible to consider changes in measurable outcomes and, within the limitations of available data, to seek to assess the nonprogram influences on these measures. As with any evaluation that seeks to identify program impacts, a major challenge was establishing whether the observed changes could be attributed to the program. To do this, one must rule out nonprogram influences. This required a plausible "counterfactual"--a description of what would have occurred in the program's absence.[1] In the analysis of crime data, as presented in chapter 3, the counterfactual for each evaluation site was the crime trend observed in the corresponding nontarget area--the rest-of-city or rest-of-county area in which Weed and Seed was not implemented. One presumes that the nontarget trend, as observed during the postimplementation periods of 1 to 5 years, reflected the pervasive underlying influences on crime. By looking at changes in each target area's crime rate relative to its corresponding nontarget area, one takes account of economic developments, demographic shifts, or anticrime initiatives unrelated to Weed and Seed. There are inherent limitations to this approach which require a careful interpretation of results. Possible overstatement of program results The extent to which the Weed and Seed effort may have geographically shifted criminal activity from a target area to other areas of the same city or county to some extent mitigates favorable conclusions drawn from a decline in the target area crime rate. Even without displacement, one would overstate the program's effects on crime if there were other simultaneous crime-reducing factors specific to the target area. The displacement of crime does, however, still reflect some increased "cost" to those engaging in criminal behavior. One should also note the Weed and Seed strategy includes the implementation of weeding and seeding activities in those areas to which crime has been displaced. Possible understatement of program results To the extent that weeding efforts (such as high-level task forces) may successfully put major drug traffickers out of business and interrupt the supply of drugs in a city, it would be incorrect to conclude--from the absence of any relative decline in the target area crime rate--that the program achieved no favorable outcome. Furthermore, selective high-level antidrug enforcement of this kind--even if highly successful at apprehending the major narcotics traffickers in the city--would typically show no discernible effect on arrests in the target area. Similarly, this analysis does not capture the extent to which youth-targeted seeding activities reduce the tendency to commit crimes later in life. It also cannot assess, except qualitatively, the upward pressure on crime rates that derives from such things as increases in a target area of gang activity, new illicit drug distribution channels, and the like. With respect to data analysis, a different approach was used to account for effects unrelated to Weed and Seed. Because any shift in the demographic composition of target area respondents might itself have shifted the pattern of survey responses between 1995 and 1997, the chapter 4 analysis focused on a specific demographic subgroup. This subgroup, nonelderly respondents (those less than 60 years old) who had resided in the target area for at least 2 years, constituted a significant minority (if not the majority) of respondents in each target area during both survey years. Together, the findings from law enforcement data and the resident survey allow an assessment of whether the evaluation sites achieved the intended changes in measurable crime-related outcomes. Although one cannot attribute these observed effects to Weed and Seed, the evidence can suggest whether the program--perhaps in conjunction with other developments--had the desired effect. The analysis involved investigating a range of measurable outcomes, taking steps to account for nonprogram influences, and examining the site-by-site pattern of intertemporal changes. Exhibit 5.1 summarizes the findings from the analysis of law enforcement data and survey data, as presented in chapters 3 and 4, respectively. For any given outcome measure, an observed change in the favorable direction is indicated by one, two, or three asterisks, depending on the strength of the supporting evidence. A blank indicates no evidence of a favorable change. An entry of "na" indicates data were not available from that site for that outcome measure. Some of the outcomes shown in exhibit 5.1 are more important than others. From the analysis of law enforcement data, the primary measure of interest is the rate of Part 1 crimes. The rates of drug arrests and Part 1 arrests are intermediate measures, reflecting more the level of law enforcement efforts than the results of such efforts. From the survey data analysis, the primary crime-related measures are those pertaining to personal safety, severity of drug crime, severity of violent crime, and victimization. (Each of these measures refers to one or more related survey questions.) Based on this summary of empirical evidence, one can group the sites into four categories, first according to the evidence of reduced Part 1 crimes and then, within each category, according to the evidence of improved public perceptions:[2] 1. Pittsburgh and Hartford exhibited strong evidence of reduced Part 1 crimes and improved public perception of crime-related measures. 2. Manatee/Sarasota (North Manatee) and Shreveport exhibited substantial evidence of reduced Part 1 crimes. 3. Seattle, Akron, Las Vegas (West Las Vegas), and Manatee/Sarasota (South Manatee) exhibited some evidence of reduced Part 1 crimes. 4. Salt Lake City and Las Vegas (Meadows Village) exhibited no evidence of reduced Part 1 crimes. Hartford and Pittsburgh stand out among the rest, with strong evidence of favorable outcomes regarding both crime and the public perception of crime. At the other extreme are Meadows Village in Las Vegas and Salt Lake City, which showed no improvement in their Part 1 crime rates relative to their corresponding nontarget areas. The remaining six target areas all exhibited either "some" or "substantial" evidence of reduced crime. Among these, the survey findings ranged from strong evidence of favorable change on multiple dimensions (North Manatee) to no evidence of favorable change on any dimensions (Shreveport). Factors favoring successful implementation of Weed and Seed The final question addressed here is this: Based on the experience of sites participating in the national evaluation, what factors appear to promote successful implementation of the program and thus promote the achievement of the program's intended results? To address this question, one must consider the site characteristics and program features that appear to distinguish the target areas as categorized according to their measurable outcomes. We identify these distinguishing factors below, with respect to community setting, programmatic design, concentration of funds, and the exercise of leadership and partnership. To identify the distinguishing features, one quickly recognizes there are no clearly defining characteristics among either the seemingly more or less effective programs. Features that appear common among more effective sites are not exclusive to such sites; the same is true for features more commonly found among less effective sites. Factors that may appear related to effective program implementation are neither necessary nor sufficient for success. For this reason, the discussion below refers to factors that appear to favor or promote successful implementation. Community setting There are preexisting features of the program setting that may make Weed and Seed easier or more difficult to operate effectively. Strength of the social and institutional infrastructure. In sites with an established network of community-based organizations and community leaders, it is often easier for a program such as Weed and Seed to find ways to build consensus about program goals and strategies. This was true in Seattle and Shreveport. The contrast was especially clear in sites with multiple target areas. In Manatee/Sarasota, the community infrastructure was better established in North Manatee than in South Manatee; in Las Vegas, it was better established in West Las Vegas than in Meadows Village. Severity of crime problems. Some types of crime are inherently more intractable than others. For example, gang violence may be so deeply wound into the social fabric of the community that progress will be slow, perhaps barely discernible over multiyear periods. The level of overall crime also seems to matter. Crime reductions may be easier to achieve in areas that have exceptionally high crime rates, but less deep-seated crime (such as Stowe Village in Hartford). Law enforcement efforts can proceed on multiple fronts, with results attainable even in the short term. Locational advantages favoring economic development. Some sites may have greater potential for economic development than others because of their proximity to commercial areas, such as the Hill District in Pittsburgh. Revitalization is easier to bring about amid a growing regional economy (Seattle) than in a community suffering from a declining industry, such as oil (Shreveport) or steel (the Hazelwood and Homestead areas of Pittsburgh). Transiency of the community population. In neighborhoods where the population is highly transient, such as Meadows Village in Las Vegas, it is more difficult for seeding programs to "take root" or for community policing to build strong police-community relations. This contrasts with more stable populations, in which residents feel more committed to neighborhood improvement, as in West Las Vegas or North Manatee. Program design The mix of weeding and seeding activities and the sequence of these components appear to be important factors in gaining community support. o Early seeding. Chapter 2 noted the importance of building community trust by implementing seeding activities at the same time as--if not before --weeding activities. Early, intensive weeding, in the absence of seeding, may stir opposition from residents who feel that a crime crackdown without any simultaneous efforts to build community will succeed only in giving a police record to neighborhood youths. This lesson was learned the hard way in Las Vegas (Meadows Village), Pittsburgh (Hill District), and Seattle. o Sustained weeding. It is misleading to think weeding can be effective as a one-time or short-lived initiative. (This misconception--and the notion that law enforcement efforts should precede human services--may be encouraged, unfortunately, by the Weed and Seed name itself.) Sites such as Pittsburgh learned that criminal activity will simply resume at its earlier levels unless law enforcement efforts are sustained. o High-level task forces, combined with community policing. Effective strategies against drug trafficking appear to require a combination of high-level interagency task forces and community-level police presence. Such efforts appeared to be successful in West Las Vegas and North Manatee. o Active prosecutorial role. This program element, viewed as an integral part of Weed and Seed, appeared to be lacking in most sites--even in some, such as Pittsburgh and Shreveport, which achieved reductions in their crime rates. Once again, West Las Vegas and North Manatee appeared to succeed more than other sites in aggressively pursuing prosecution at both the local and Federal levels. Concentration of funds Sites appeared to have greater success if they concentrated program resources on smaller populations, especially if they could similarly channel other public funds and leverage private funds. Funding "intensity." Not surprisingly, the more narrowly defined the target population, the greater the chance of success in fighting crime with a constrained program budget. Among the evaluation sites, Hartford used the greatest funding intensity, applying more than $2.4 million to a community (Stowe Village) of only 1,300 residents living in 0.1 square mile. This contrasted sharply with Salt Lake City, where only $550,000 was committed to a target area of 22,000 residents in 6.3 square miles. Channeling and leveraging other funds. Also, as would be expected, sites that commanded other funding, either by channeling public funds to the target area or by securing private funding, mounted more effective programs. This was especially true among sites where Weed and Seed funding intensity was low. Las Vegas made considerable use of both Federal and State program funds in West Las Vegas, as did Pittsburgh in the Hill District. Manatee County and Seattle leveraged substantial amounts of private funding. Leadership and partnership A less tangible ingredient that seemed to characterize the more successful programs was the active and constructive leadership of key individuals. Depending on the site, these key actors occupied different institutional roles; they included community leaders, program staff, police personnel, mayors, FBI Special Agents in Charge, and U.S. Attorneys. In all instances, however, they adopted a style of leadership that encouraged cooperation and coordination. In sites that struggled to implement the program, individuals in these key roles adopted a more confrontational style that bred conflict and divisiveness. By its very nature, Weed and Seed places a great premium on effective coordination of groups with different organizational missions and who respond to different constituencies. To establish effective relationships among these organizations required personal energy and initiative. Leadership meant persuading other partners to lay aside jurisdictional "turf" issues or perhaps even personal rivalries. From their own early experiences, a number of sites learned the importance of adopting a "bottom-up" approach to identifying problems and proposing solutions. A participatory process involving community leaders and residents--although more time consuming--typically produced decisions that were more easily implemented and which had a greater chance of success. A more heavy-handed, top-down approach, in contrast, tended to break down. Once disagreements arose, it became difficult to bring the key players back to the table. Finally, implementation strategies that relied on bottom-up approaches were most effective when combined with efforts to build capacity among participating local organizations. This required a longer-term perspective about the program and its potential to bring about community change. However, such sites--including some that achieved substantial crime reductions within the time period analyzed--have established a stronger foundation and more sustainable basis for further community-targeted initiatives. Notes 1. The most rigorous means of establishing the counterfactual is an experimental design in which the intervention--in this case Weed and Seed--is not implemented among a randomly selected set of subjects, in this case, sites. Such a design was infeasible in this context. Another approach would have been to match each evaluation site with a comparison area of similar baseline characteristics. At an earlier stage of this evaluation, NIJ and the Institute for Social Analysis identified such comparison sites and included them in the 1995 survey of community residents and other baseline data collections efforts. Soon thereafter, however, NIJ and Abt Associates concluded that the comparison sites were not sufficiently comparable with their corresponding target areas in demographic characteristics and crime trends. The matched site approach was not pursued further. See Dunworth, Terrence, et al., "Overview of Institute for Social Analysis National Evaluation Baseline Data and Implications of the Data for the Weed and Seed Impact Evaluation," Abt Associates Inc., Cambridge, Massachusetts, 1996. 2. Meadows Village in Las Vegas and South Manatee in Manatee/Sarasota, where the survey was not conducted, are considered to have indeterminate evidence on improved public perceptions and are listed within their group below the heading of sites for which the survey provided some evidence of a favorable change in resident perceptions. ------------------------------- Chapter 6 Federal funding for Weed and Seed is modest when compared to other intervention programs and is only increasing slowly. Moreover, the number of Weed and Seed sites is expanding rapidly. Assuring a successful future for Weed and Seed therefore demands that policymakers make the most efficient possible choices when awarding funds. Evaluation results show that locations with clearly defined goals, small target areas, and a high degree of resident support are most likely to have a beneficial effect on crime rates. They also suggest that Weed and Seed has served as a strong stimulant to community mobilization and interorganizational cooperation and collaboration. Therefore, program focus and community mobilization--program cornerstones--should be key in determining program expansion. Policy Implications Policymakers now face a number of strategic choices as they chart the future direction of Weed and Seed. These choices revolve around the issues of designating sites for continued funding, selecting sites for new awards, and allocating funds among participating sites over time. The central policy question is how to use program funds most effectively in order to make the greatest long-term contribution to the program's goals of controlling crime and promoting a safe living environment for city residents. The findings presented in this report indicate significant favorable effects of Weed and Seed on key outcome measures for some sites and for some time periods. The evidence is modest in terms of statistical significance, but the indicators consistently point in the favorable direction. It is important to recognize that the standards of evidence applied in this report--although fully consistent with those normally applied in evaluation research--are considerably more strict than policymakers might consider appropriate. Given the need for policymakers to consider the program's future and their need to base decisions on whatever information is available, it is important to speak to these policy issues as definitively as possible. Program effectiveness and community mobilization The experience of the eight sites studied in the evaluation suggests that Weed and Seed affected the target areas through either (or both) of two avenues. The first, here called "program effects," relates to specific initiatives that focus on law enforcement and crime prevention. These activities, on both the weeding and seeding side, appear to have had varying degrees of success in reducing crime and improving perceptions of public safety, within the 2- to 5-year postimplementation periods analyzed here. As noted previously, the degree of success seems related to the community setting, program design, and concentration of funds. The second avenue of change in these sites has been the process of community mobilization. In many sites, Weed and Seed provided a catalyst for greater involvement of neighborhood residents and community-based organizations in community revitalization efforts. Again, some sites were more predisposed to participatory problem-solving arrangements than others, by virtue of their preexisting infrastructure and active leadership by both individuals and established organizations. Such factors enabled these sites to overcome the natural barriers to coordination among local interest groups and public agencies. This approach has long been a cornerstone of the Weed and Seed philosophy, but it has been deliberately and strategically emphasized in recent years. The consequence has been a significant increase in community mobilization. These two mechanisms of change are clearly interrelated. To the extent that program initiatives require organizational linkages and resident support, such initiatives cannot be launched if the relevant actors have not been mobilized. Similarly, the mobilization process is facilitated by first having demonstrated some effectiveness of early program initiatives. Sites differ in their emphasis on program initiatives (in the interest of program effectiveness) and capacity-building or infrastructure development (in the interest of community mobilization). These design choices are somewhat like consumption-investment decisions. A series of program initiatives may yield significant short-term results, which cumulate over time--but where the effects may endure only as long as the funding. By contrast, resources devoted to capacity-building may not yield any effects for a substantial time period but may establish a foundation for sustainable, long-term progress. It is useful to consider the interplay of these two forces--program effectiveness and community mobilization--in sites of varying size, assuming (as has been the practice) that Weed and Seed funding is provided in annual grants that do not vary by size of site. In larger sites, program effects are typically more difficult to achieve because the Weed and Seed funding will constitute a smaller share of total public and private resources devoted to reducing and preventing crime. In the extreme, Weed and Seed may represent the proverbial drop in the bucket. In smaller target areas, such as the Hartford site, a significant effect is more attainable because Weed and Seed facilitates a meaningful increase in weeding- and seeding-type activities. Selecting sites and allocating funds The growth in the number of sites receiving Weed and Seed funds each year has far exceeded the growth in annual program funding. One result is that demonstration sites, which previously received as much as $750,000 per year, now typically receive $250,000 per year. What do the findings imply about the way policymakers should approach selecting sites and allocating program funds? There are three major implications of this research. It is important to note that the following points do not address the issue of whether to increase or decrease the overall level of national funding. Rather, they address the issue of how best to allocate funds. First, given the importance of institutional capacity-building and infrastructure development and the increasing emphasis placed on these factors by EOWS, Weed and Seed should consider providing additional technical assistance to funded sites. The lessons learned from the more successful sites--and the less successful ones--on capacity-building and infrastructure issues can be generalized to a large degree, and new sites should receive the benefit of this experience. This seems especially important to promote the type of partnership arrangements that exist in the more successful programs examined. Second, in the interest of maximizing program effectiveness and community mobilization, Weed and Seed should seek the highest feasible geographic concentration of funds in the program sites. Given congressional funding constraints, this implies use of a highly selective process in choosing sites to receive new awards and/or shortening the number of years that ongoing sites receive program funding. Thus, in selecting sites, Weed and Seed should place its funding priority on sites with geographically small target areas and with favorable community settings and program designs--i.e., with features favoring successful implementation. ------------------------------- About the National Institute of Justice The National Institute of Justice (NIJ), a component of the Office of Justice Programs, is the research agency of the U.S. Department of Justice. Created by the Omnibus Crime Control and Safe Streets Act of 1968, as amended, NIJ is authorized to support research, evaluation, and demonstration programs, development of technology, and both national and international information dissemination. Specific mandates of the Act direct NIJ to: o Sponsor special projects, and research and development programs, that will improve and strengthen the criminal justice system and reduce or prevent crime. o Conduct national demonstration projects that employ innovative or promising approaches for improving criminal justice. o Develop new technologies to fight crime and improve criminal justice. o Evaluate the effectiveness of criminal justice programs and identify programs that promise to be successful if continued or repeated. o Recommend actions that can be taken by Federal, State, and local governments as well as by private organizations to improve criminal justice. o Carry out research on criminal behavior. o Develop new methods of crime prevention and reduction of crime and delinquency. In recent years, NIJ has greatly expanded its initiatives, the result of the Violent Crime Control and Law Enforcement Act of 1994 (the Crime Act), partnerships with other Federal agencies and private foundations, advances in technology, and a new international focus. Some examples of these new initiatives: o New research and evaluation are exploring key issues in community policing, violence against women, sentencing reforms, and specialized courts such as drug courts. o Dual-use technologies are being developed to support national defense and local law enforcement needs. o The causes, treatment, and prevention of violence against women and violence within the family are being investigated in cooperation with several agencies of the U.S. Department of Health and Human Services. o NIJ's links with the international community are being strengthened through membership in the United Nations network of criminological institutes; participation in developing the U.N. Criminal Justice Information Network; initiation of UNOJUST (U.N. Online Justice Clearinghouse), which electronically links the institutes to the U.N. network; and establishment of an NIJ International Center. o The NIJ-administered criminal justice information clearinghouse, the world's largest, has improved its online capability. o The Institute's Drug Use Forecasting (DUF) program has been expanded and enhanced. Renamed ADAM (Arrestee Drug Abuse Monitoring), the program will increase the number of drug-testing sites, and its role as a "platform" for studying drug-related crime will grow. o NIJ's new Crime Mapping Research Center will provide training in computer mapping technology, collect and archive geocoded crime data, and develop analytic software. o The Institute's program of intramural research has been expanded and enhanced. The Institute Director, who is appointed by the President and confirmed by the Senate, establishes the Institute's objectives, guided by the priorities of the Office of Justice Programs, the Department of Justice, and the needs of the criminal justice field. The Institute actively solicits the views of criminal justice professionals and researchers in the continuing search for answers that inform public policymaking in crime and justice.