U.S. Department of Justice Office of Justice Programs National Institute of Justice National Institute of Justice R e s e a r c h i n B r i e f March 2001 Issues and Findings Crime Mapping and Analysis Discussed in this Brief: An as- sessment of how community orga- by Community Organizations nizations in Hartford, Connecticut, used the Neighborhood Problem Solving (NPS) system, a computer- in Hartford, Connecticut based mapping and crime analysis technology. The NPS system en- By Thomas Rich abled users to create a variety of maps and other reports depicting Crime mapping has become increasingly community policing field offices. Devel- crime conditions by accessing a popular among law enforcement agencies oped with NIJ funding, the system enables database containing the most re- and has enjoyed high visibility at the Fed- users to create a variety of maps and other cent 2 years of incident-level police information. eral level, in the media, and among the reports depicting crime conditions by us- largest police departments in the Nation, ing a database containing the most recent Key issues: The objective in Hart- most notably those in Chicago and New 2 years of incident-level police informa- ford was to extend basic mapping York City. However, there have been few tion, including citizen-initiated calls for and crime analysis technologies efforts to make crime mapping capabilities service, reported crimes, and arrests. Al- beyond the law enforcement com- available to community residents and or- though some police departments routinely munity by putting them into the ganizations, although a National Institute publish aggregate crime statistics and a hands of neighborhood-based organizations so they could analyze of Justice (NIJ)-funded project in Chicago few publish incident-level crime informa- incident-level data and produce in the late 1980s aimed at introducing tion on the Internet, the objective in Hart- their own maps and reports. The mapping to community groups showed ford was to extend basic mapping and assessment discussed in this Brief that this technology could benefit them.1 crime analysis technologies beyond the was conducted to determine the Access to timely and complete incident- law enforcement community so that extent to which community organi- level crime information—among other neighborhood-based organizations could zations used the system, to under- obstacles—has inhibited the spread of analyze incident-level data. stand how these organizations mapping to communities. used the software, and to assess NIJ sponsored an assessment of Hartford’s the effect of the system on com- To get the community more involved in system by Abt Associates Inc. to deter- munity organization effectiveness, crime prevention, Hartford, a city of mine the extent to which community orga- perceptions of neighborhood 130,000 in central Connecticut, expressed nizations used it, to understand how these safety and quality of life, and an interest in providing community-based organizations used the software, and to as- police-community relations. organizations with a crime mapping and sess the effect of the system on community Key findings: analysis system. This technology—dubbed organization effectiveness, perceptions of the Neighborhood Problem Solving (NPS) neighborhood safety and quality of life, ● Seven of the 14 sites used the NPS system regularly. Two sites system—was developed and implemented and police-community relations. This used the system on several in 1997 and 1998 in 14 locations through- Research in Brief discusses the findings occasions but not on a regular out Hartford, including community organi- of this research. basis. The remaining five sites zation headquarters, public libraries, and either did not use the system at all or used it only once. Support for this research was provided through a transfer of funds to NIJ

continued… from the Office of Community Oriented Policing Services. COMMUNITY ORIENTED POLICING SERVICES U.S. DEPARTMENT OF JUSTICE R e s e a r c h i n B r i e f

Issues and Findings Project origin strategies, and assessing results. City . . . continued employees were trained in community- In November 1994, Hartford was 1 of 12 oriented government techniques. In addi- ● Factors explaining the variety of cities to receive a Comprehensive Com- tion, CCP began funding the community use patterns include the community munities Program (CCP) grant from the organization United Connecticut Action for organization leader’s degree of U.S. Department of Justice’s Bureau of Neighborhoods in 1997 to provide addi- interest in the system, available Justice Assistance. The Hartford CCP tional training and technical assistance to personnel resources at the site, the sought to shift the focus of city govern- site’s experience with community the committees. ment away from the professional civil ser- organizing and community-based problem solving, and the relevance vice model to a decentralized government Beginning in 1995, the committees were of both the system and the crime that recognizes residents as its “custom- eligible to receive personal computers data to the organization’s primary ers” and attends to problems and solu- equipped with standard business produc- work objectives. tions they identify. In 1988, the police tivity software and a laser printer. Many department had initiated community- committees immediately took advantage ● While all regular-use sites made oriented policing with the formation of the of this offer, although some did not apply extensive use of crime data, some Community Service Officer (CSO) unit. sites placed the highest value on for their computers until 2 or 3 years CSOs focused exclusively on problem arrest data and others closely exam- later. The NPS system was eventually in- ined calls-for-service data. Maps solving within a specific neighborhood. stalled on these computers. The Hartford were the most popular report The department also formed partnerships Police Department and Abt Associates format. with other agencies, including schools, jointly administered the NPS program. youth organizations, and other criminal The police department provided data on ● The majority of frequent NPS us- justice agencies. In 1995 and 1996, calls for service, crimes, and arrests to ers felt they gained a “much better Hartford established a number of CCP- Abt Associates, which formatted the data understanding” of crime conditions supported initiatives and programs that in their neighborhood using the sys- and provided it to the community organi- tem. The majority of frequent users created, in the words of one veteran com- zations. Abt Associates maintained the believed that access to comprehen- munity organizer, “an infrastructure” that NPS system software, while the city of sive crime data did not change the supported the NPS system. According to Hartford maintained the computers. extent to which they felt safe in the this organizer, having this infrastructure Trinity College in Hartford provided the neighborhood. in place was a necessary condition for committees with e-mail and Internet widespread acceptance and use of the accounts beginning in 1996. ● Hartford’s extensive problem- NPS system. solving infrastructure, including neighborhood problem-solving Citizen-based problem-solving committees, System capabilities committees, veteran community one of the most important CCP initiatives, In 1997, representatives of the problem- organizers, and a supportive envi- were formed in each of Hartford’s 17 ronment in the city government and solving committees, Hartford city officials police department, was key to the neighborhoods. In many neighborhoods, involved in the city’s CCP, Abt Associates program’s success. Without this in- these committees consisted of residents al- staff, and representatives from the Hartford frastructure, the system likely would ready active in the existing neighborhood Police Department and other city service have been used by interested par- citizen groups, but the committees were agencies held a series of meetings to deter- ties to examine crime information an entirely new group for some neighbor- mine what information—particularly auto- but not necessarily for effecting hoods. The committees were charged with mated information—would be useful (and neighborhood change. identifying and prioritizing neighborhood could be provided) to the committees to problems and, with the help of city agen- Target audience: City- and county- support their problem-solving activities. level police, planning and administra- cies, implementing solutions to those prob- Ideally, an information system supporting tive officials, neighborhood-based lems. The city retained an independent problem solving—especially one with organizations, and researchers. consultant to conduct problem-solving mapping capabilities—would contain data training for the committees. The training from a variety of sources, including city, focused on recognizing and analyzing State, and Federal agencies; private and problems, developing and implementing nonprofit organizations; and neighborhood-

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

based organizations. Such data would vanced computer-aided dispatch, case for service, crimes, and arrests. No enable maps to show a multidimen- incident reporting, and online arrest names or other personal identifying sional view of crime. For example, one booking systems. The department has information are provided (see “Data layer of a map display could contain a used computerized crime mapping Elements Provided to Community descriptive variable, such as the loca- since the early 1990s, when it partici- Organizations”). Community organiza- tion of burglaries in the past month, pated in NIJ’s Drug Market Analysis tions are provided with the most recent while other layers could contain pos- Program. Since the mid-1990s, depart- 2 years of these data occurring in sible explanatory variables, such as ment analysts have distributed weekly the organization’s area of concern, demographic and social indicators or crime maps to command staff. which typically corresponds to one of the locations of abandoned buildings Hartford’s 17 neighborhoods. For an or liquor establishments.2 One important difference between a average-sized community organization, crime mapping tool developed for in- the 2-year timespan of data translates Committee representatives indicated ternal police use and one developed into about 12,000 calls for service, they wanted access to a variety of for use by community groups or the 7,000 crimes, and 1,000 arrests. Dur- information, including police, hous- general public involves the issue of ing the study period, Abt Associates ing, and code enforcement data that which data to make available to users. staff provided data to the community could be tied to specific addresses. The NPS system provides basic “what, organizations every 2 weeks via e-mail Unfortunately, with the exception of when, and where” information on calls police calls for service, crime, and arrest data, information was either not automated or out of date by sev- eral months, although projects were Data Elements Provided to Community Organizations under way to improve the quality of the data. By contrast, police calls for A ny police department publishing type and category, block where the crime service and arrest records were en- incident-specific information on its Web occurred, and latitude and longitude of tered in real time (i.e., while the call site or other media must address the is- where the crime occurred. taker talks to the person calling 911 sue of what specific data elements, par- ticularly those related to location, should ● For each person arrested: case num- and the suspect goes through the ar- ber; date and time of arrest; sex, age, rest booking process), and reported be made available to the public. Police departments must balance the public’s and race of the person arrested; charge crime information was typically auto- right to know against the right to privacy type and category; street address where mated about 1 week after the police of complainants, victims, and suspects. the arrest occurred; and latitude and lon- officer submitted his or her com- This important issue is the focus of a gitude of where the arrest occurred. pleted crime report. forthcoming NIJ publication. The specificity of the location information In light of these findings, CCP staff, the In Hartford, community organizations re- varied. For arrests, the street number and police department, and Abt Associates ceived incident-specific information ob- name (e.g., 542 Main Street) are pro- staff decided that the NPS system tained with the Neighborhood Problem vided, whereas only the street name and block (e.g., 500 block of Main Street) are would initially focus on providing the Solving (NPS) system, which contained the following data elements: provided for calls for service and crimes. committees with access to police calls However, geographic coordinates (lati- for service, crime, and arrest data. ● For each citizen-initiated call for ser- tude and longitude) of the street number vice: call number, date and time the call and name are included in the NPS data- Data issues was received, call-for-service type and base for calls for service, crimes, and category, associated incident report num- arrests. The coordinates coincide with Prior to participating in this project, ber (if any), block where the police unit the Hartford Police Department’s street the Hartford Police Department had was sent, and latitude and longitude of centerline file. Thus, NPS maps show the a history of deploying sophisticated where the police unit was sent. approximate point along a street where technologies in its crimefighting and the event occurred but do not indicate public safety efforts. Department staff ● For each reported crime: case number, on which side of the street the event had designed and programmed ad- date and time the crime occurred, crime occurred.

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

or U.S. mail. Because of the 2-week For all reports, users can specify a Hartford’s Part I4 crime rate in 1997 delivery cycle—and because the po- desired date and time range. Standard was nearly twice the national average, lice department entered calls for ser- options are available (e.g., for dates, the even though the city, like many other vice and arrest information in real time most recent 2 weeks, the most recent cities in the Nation, experienced a and crime information within a week of month, and the most recent year), or the significant decrease in crime in the when the officer completed the crime user can specify any desired date and 1990s; from 1987 to 1992, the city report—the data were roughly 10 days time range. Maps and graphs are pro- recorded between 20,000 and 22,000 to 2 weeks old when delivered. duced for a specified group of calls for Part I crimes per year, and by 1997, service, crimes, or arrests (e.g., narcot- the annual total had dropped to 3 The NPS software can produce five ics arrests, violent crime-related calls slightly more than 12,000. Crime, different types of reports, each of for service, or burglary crimes). Cur- however, is by no means the only issue which can be based on either calls for rently, only one group can be mapped of concern to these organizations. An service, reported crimes, or arrests. at a time. Detail lists can be sorted by organizer in one neighborhood indi- date and time, by type of event, or by cated that the four issues discussed ● A detail list shows records that meet a specified criterion (e.g., all address. In addition, detail lists can most at neighborhood meetings are (in burglaries in a neighborhood over show all events (e.g., all arrests) or only order of importance) taxes, education, the past year). events in a certain group (e.g., only nar- crime, and litter. cotics arrests). ● A top 10 list shows the most fre- Although all community organizations quent types of events (e.g., the 10 From June 1998 to March 1999 (the with the NPS system share major crime most frequently reported types of end of the evaluation period), the NPS concerns, they differ in their overall ap- crime in a neighborhood over the software was installed at 14 community proach to addressing these problems in past month). locations in Hartford: 8 community their neighborhoods. For example, organization headquarters, 4 commu- some focus on identifying issues of con- ● A time trend graph shows the nity policing substations, and 2 public cern to neighborhood residents and de- number of events occurring over a libraries. veloping approaches to address those recent set of time periods (e.g., the concerns. Other organizations focus on number of burglaries in a neigh- Site characteristics delivering services to neighborhood borhood by week over the past 6 residents, such as sponsorship of job months). The community organizations that fairs or street-lighting projects. Many of received the NPS system and were ● An event trend graph shows the the Hartford problem-solving commit- percentage change in the number trained in its use vary considerably tees work directly with police CSOs to of different types of events over a in size, resources, and experience in address specific crime problems. recent period (e.g., the percentage community organizing and problem solving. One has approximately 15 change in the number of each crime System users group in the most recent month full-time employees, most of whom are compared with the previous month). neighborhood organizers assigned to There was at least one targeted user— one of nine neighborhoods in Hartford. a person trained in the use of the sys- ● A pin map shows the location of Others have one or two full-time or tem who received and processed crime events over a specified period (e.g., part-time paid staff. By contrast, the data updates and whose reaction to the the location of all burglaries in a problem-solving committees are essen- system was monitored—at each instal- neighborhood over the past month). tially groups of concerned citizens lation site. In many cases, these tar- Although the report is referred to as with other full-time occupations and geted users showed others in their a pin map, it is actually a graduated have no paid staff or formal office organizations how to use the system. symbol map in which the size of the space. Independent of their size and Having a targeted set of users distin- icon depicting the location is pro- resources, the community organiza- guished this project from one that portional to the number of events tions and problem-solving committees involved the installation of an NPS- at that location. all appear to possess a strong concern like system at public kiosks or on the about crime and public safety. Internet to provide information to the

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

general public. In general, the system sions but not on a regular basis dur- the site’s primary objective was to pro- in this study was designed for commu- ing the data collection period. vide job training or other services to nity organizations, and there was no residents. Limited accessibility to the ● One-time users—four sites used attempt to encourage its use by the computers equipped with the NPS sys- the NPS system once during the general public. tem inhibited system use in two sites. data collection period.

The four types of targeted users were: ● Nonusers—one site did not use The types of maps and reports pro- the NPS system at all during the duced with the NPS system provide ● Block watch organizers, who data collection period. insight into the general types of police work for a community organization data and data formats community orga- and are charged with establishing At the conclusion of the initial training nizations in other jurisdictions might new block watches and supporting session at each installation site, all find particularly useful. Each NPS run existing block watches. The city of targeted users reacted enthusiastically was based on calls for service, crimes, Hartford provided funding to com- to the system, praising its ease of use or arrests (see exhibit 1). Significant munity organizations to hire these and listing ways in which their organi- site-by-site variations exist, however, individuals. zation could use the system. Yet, 5 of among the regular users of the system. the 14 sites either did not use the sys- ● Neighborhood organizers, who For example, three regular NPS users tem at all or used it only once. Why work at community organizations showed little interest in calls-for-service did some community organizations use and undertake community organizing data; their sites used those data in less the system and not others? activities in the neighborhood to than 10 percent of their runs. Users at which they are assigned. In particu- The primary reason some sites regu- another regular-use site, however, lar, neighborhood organizers support larly used the system was that organi- were most interested in calls for ser- the problem-solving committees. zation leaders affiliated with the site vice. Users at a third regular-use site took a personal interest in it. In con- focused most often on arrest data, re- ● Community organization staff, flecting a strong interest in monitoring which refers to other individuals at trast, the targeted users in several other neighborhoods were largely on the police response to neighborhood community organizations who have problems. a variety of responsibilities, includ- their own and without supervisors to ing administration, management, encourage them to use the system. In Four of the seven regular-use sites grant writing, and community addition, several users were relatively preferred the pin map report format. organizing. new to the field of community organiz- Across all sites, the pin map was the ing and naturally had a more difficult most preferred report format, and it ● Problem-solving committee time integrating an information tool was used in 39 percent of all NPS members, who have other full- such as the NPS system into their time occupations and positions. runs. The other report formats also organizing activities. showed significant use, ranging from System utilization Personnel resources also influenced 23 percent (the detail list) to 8 percent use patterns. Targeted users at many (the top 10 list) of all runs, thus rein- The most basic evaluation issue for sites where the system was used regu- forcing the notion that maps are only the NPS system was how frequently larly were persons whose full-time job one method of analyzing police data the system was used. Across NPS was community organizing. At some and that a variety of report formats is sites, the frequency of system usage sites where the system saw little use, required to provide useful information varied. however, targeted users had other full- to users of police data (see exhibit 2). ● Regular users—7 of the 14 sites time occupations and did not have ac- Although information about the pre- used the NPS system regularly on cess to volunteers, interns, or full-time ferred data types and report formats a monthly or more frequent basis organizers who could use the system. used in Hartford may provide insight during the data collection period.5 Some organizations did not regularly into the possible preferences of com- use the system because it was not par- ● Irregular users—two sites used munity organizations in other cities, ticularly relevant to the targeted user’s the NPS system on several occa- the specific type of data displayed in primary work objectives—for example,

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

the NPS system was useful for getting Exhibit 1. Percentage of reports generated, by type of data people to attend block watch meetings. The block watch organizer in a neighbor- Crime hood perceived as one of Hartford’s safest 43% commented that the maps and reports she showed to potential block watch members helped convince them that “there is real crime in their quiet neighborhood, so a block watch would be beneficial.” Arrest 29% Supporting existing block watches. In several neighborhoods, block watch organizers regularly produced maps and other reports and shared them with attendees at block watch meetings. Calls for service 23% Organizers consistently commented that the maps and reports served as conversation starters for the meetings Note: n = 2,444. and provided valuable information. In one neighborhood, support for the the maps and reports is likely to vary Just as the sites used the NPS system to block watches was highly systemized depending on the issues of importance varying degrees, they used it in a variety because community organization staff, to users in those cities. Data prefer- of ways and for a variety of purposes. together with several block watch lead- ences in Hartford may be representa- Primary uses of the maps and reports ers, designed and distributed a biweekly tive of other urban areas, however. include the activities described below. “CrimeData Update” containing NPS Narcotics-related information was the reports to all block watch leaders. most common type of data to appear in Forming block watches. Veteran Hartford sites’ NPS pin maps, followed block watch organizers commented that closely by Part I crime-related infor- mation (see exhibit 3). Together, these Exhibit 2. Percentage of reports generated, by report format data constituted nearly 80 percent of Top 10 list all NPS pin maps. The most common 8% Part I crime-related maps across all sites were burglaries (14 percent of all Time trend graph pin maps), followed by auto theft and 11% aggravated assaults (9 percent each). Pin map 39% Overall, the interest in Hartford was serious crime, as opposed to less seri- ous crime and quality-of-life issues. Nevertheless, there were significant variations across the regular-use sites. Event trend graph 19% The percentage of maps depicting narcotics activity, for example, ranged from a low of 8 percent to a high of 81 percent. The percentage of maps depicting noncriminal activity (e.g., Detail list accidents, motor vehicle infractions) 23% ranged from 0 to as much as 24 percent. Note: n = 2,444.

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

Exhibit 3. Types of data appearing in Hartford NPS pin maps Group Major Category Number of Runs % of Total Cumulative % Narcotics Narcotics Related 411 43 43 Burglary Part I Crime 135 14 57 Auto Theft Part I Crime 89 9 66 Aggravated Assault Part I Crime 83 9 75 Breach of Peace Part II Crime 35 4 78 Other Part II Crimes Part II Crime 17 2 80 Larceny Part I Crime 17 2 82 Prostitution Part II Crime 14 1 83 Driving Laws Part II Crime 13 1 85 Simple Assault Part II Crime 11 1 86 General Violent Crime Part I Crime 11 1 87 General Property Crime Part I Crime 11 1 88 Robbery Part I Crime 10 1 89 Domestic Disputes Part II Crime 8 1 90 Minor Crime Against Property Part II Crime 7 1 91 Alarms Part II Crime 7 1 92 Property Damage Accident Noncrime 6 1 92 Personal Injury Accident Noncrime 6 1 93 Motor Vehicle Laws Noncrime 6 1 93 Liquor Laws Minor Infraction 6 1 94 Traffic Details Noncrime 5 1 95 Intoxication Part II Crime 5 1 95 City Ordinances Minor Infraction 5 1 96 Administrative Noncrime 5 1 96 Weapons Part II Crime 4 0 97 Sex Offenses Part II Crime 3 0 97 Receiving Stolen Property Part II Crime 3 0 97 Rape Part I Crime 3 0 98 Gambling Part II Crime 3 0 98 Fire Related Noncrime 3 0 98 Drunk Driving Part II Crime 3 0 98 Community Assistance Noncrime 3 0 99 Animal Complaint Noncrime 3 0 99 Homicide Part I Crime 2 0 99 Forgery Part II Crime 2 0 99 Offenses Against Family Part II Crime 1 0 100 Noninvestigated Accident Noncrime 1 0 100 Miscellaneous Noncrime Noncrime 1 0 100 Minor Juvenile Cases Part II Crime 1 0 100 Minor Part II Crime Part II Crime 1 0 100 Totals 960 100

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

Supporting neighborhood-level officials to substantiate requests for resources to combat the problem. The organizations. Members of citizen action in a hot spot or for additional leader of one Hartford community orga- groups, neighborhood organizers, and police resources. nization indicated that the primary other community organization staff value of the NPS system was using data used the NPS system to provide infor- Problem-solving and crime to highlight the chronic problems in a mation to neighborhood-level organi- prevention applications neighborhood. He added that his orga- zations, much like a police crime nization was often “somewhat skittish analyst produces a set of maps for a Targeted users and the people for whom about going after a problem if the only police command staff meeting. maps and other reports were produced evidence is citizen perceptions.” used the NPS system to accomplish a Raising neighborhood awareness. variety of specific or general objectives Although in most cases the NPS system One community organization used (see “Study Methodology”). One of the was used to confirm and quantify known the NPS system to disseminate crime most common objectives related to problems, targeted users noted that in information to residents throughout the the NPS system was identifying and some instances the system highlighted a neighborhood on National Night Out, quantifying crime hot spots, including previously unknown problem (at least to an annual nationwide event in which specific addresses, streets, and sections the community group or neighborhood residents participate in outdoor activi- of neighborhoods. In most cases, the problem-solving committee). This find- ties designed to mark progress in “tak- quantifying component was signifi- ing indicates what some in the Hartford ing back the night” from drug dealers cantly more important than the identify- Police Department hoped would be an and other criminals. Community orga- ing component because community outcome of this project—that citizens nization staff used the NPS software organizers and other persons active in would become effective crime analysts. to produce a two-page crime summary a neighborhood were already aware of handout and delivered it to neighbor- the most problematic locations in their Whether a property was a known or an hood block-party hosts. area. What those individuals ordi- unknown problem, community organiza- narily could not do was quantify the tions used several techniques to target Distributing information to law seriousness of the hot spot in terms of specific problem properties, including enforcement officials. Several com- citizen complaints, reported crimes, discussing the problem with the police munity organizations used NPS maps and arrests. CSO and other police officials, talking and reports to facilitate communica- to the owners, and publicizing the prob- tion with police and prosecution Quantifying the seriousness of a hot lem in the media. In July 1998, the officials. CSOs, for example, regularly spot serves two key purposes for com- Connecticut Legislature provided an ad- attended neighborhood problem- munity groups and neighborhood resi- ditional tool that the police, prosecutors, solving committee meetings and were dents. First, maps and reports confirm and community organizations could use typically called on to explain and in- the organizers’ or residents’ perception to target specific properties. The Nui- terpret the NPS output. During some of the problem. A number of targeted sance Abatement and Quality of Life of these discussions between residents users indicated that the NPS output Act gives the State, through the State’s and police officials, residents com- served as a “reality check” and let Attorney’s office, the right to bring a civil mented on apparent inaccuracies in residents know the problems they saw action against any property owner whose the police data, highlighting what one were, in fact, real. Conversely, the property creates a public nuisance. The veteran community organizer believed NPS system was also used to suggest law defines a public nuisance as three was the most important use of the NPS that a problem thought to be serious or more arrests (on different dates) for a system: improving the exchange of in- was, in fact, not that serious. The sec- variety of crimes, including prostitution formation between the community and ond and more important purpose for and the sale or possession of narcotics. the police and, in particular, ensuring quantifying problems is that commu- The value of the NPS system in devel- that the police and the community nity groups, armed with hard evidence oping cases for the Nuisance Abatement have a common understanding of of a problem, can make a stronger case Act is obvious, and several community neighborhood problems. In other sit- both to neighborhood residents to get organizations in Hartford used the sys- uations, the maps and reports were involved and to the police and other tem to identify properties covered under used at meetings with law enforcement city agencies to provide additional the Act. The organizations then notified

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

the Connecticut attorney general’s of- tions and encourage residents to focus quently they reported it to the police. fice, attaching NPS output to their let- on crime prevention. As one neighbor- As a result, the organization mounted ters as supporting documentation. hood organizer said, “The main use of a campaign (using NPS reports that the system is to bring the neighbor- showed how few citizen drug-related A number of community organization hood together—getting [neighbors] to calls for service were made in the leaders in Hartford said they believed focus on an issue and mobilize around neighborhood) to encourage residents a primary use of the NPS system would it.” One community organization noted to contact the police. The number of be to raise awareness throughout the how frequently residents complained calls for service increased substan- neighborhoods regarding crime condi- about drug activity but how infre- tially following this outreach effort

Study Methodology T he evaluation of the Hartford NPS such a feature would pose an obstacle to and Abt Associates staff were informed system had three primary objectives: system use. of only a few instances when persons not affiliated with a community organi- ● Use level—to determine system uti- To assess how the committees used the zation or citizen group used the NPS sys- lization, including measuring how often system and its effect on users, Abt Associ- tem. As such, this evaluation focused on community organizations used the sys- ates staff conducted informal interviews how community organizations, not the gen- tem and what types of maps and re- with and a formal survey of system users eral public, use crime information. ports they created. and attended neighborhood meetings at which crime issues were discussed. The At each site, Abt Associates staff installed ● Use purpose—to understand what survey focused on such issues as how us- the software, provided training to one or the organizations did with the maps ers decided what types of maps and re- more persons, made general suggestions and reports they produced and what ports to produce, how often users referred on how the software might be used, they hoped to accomplish with them. to the maps and reports, how the maps and provided ongoing technical support and reports affected users’ understanding to users both in person and over the ● Use effect—to assess how use of of crime conditions and their feelings telephone. However, because the re- the system affected perceptions of about neighborhood safety and quality of searchers wanted to find out what orga- neighborhood safety and quality of life, life, whether the maps and reports changed nizations would do with the system, community organization effectiveness, users’ attitudes toward the police, how Abt Associates did not develop a plan for and police-community relations. often users shared maps and reports with how organizations should use the system others, and how the maps and reports or help them integrate the system into The first objective, measuring system use could have been made more useful. their workflow. Thus, while Abt Associ- levels, is perhaps the most fundamental ates was by no means a disinterested ob- question in any evaluation of an informa- Identifying NPS system users for the evalu- server of the system, the researchers also tion tool. Special purpose software, espe- ation was not difficult because the soft- were not the driving force behind its use. cially when not integrated with the user’s ware ran on stand-alone PCs and, for this workflow patterns, often falls into disuse project, only Abt Associates staff had the after an initial period of use. The research software installation disks.a Identifying a. However, NPS software is now available from team fully expected this at some of the users would be more difficult if one were NIJ. installation sites. For the NPS evaluation, examining an Internet-based police infor- b. Some NPS installation sites, such as public researchers relied primarily on tracking mation sharing system. Although the gen- libraries, are open to the public; others are in logs that recorded information about eral public in Hartford could walk in and community police field offices, which are each map or report that users produced. use some of the computers with the NPS generally locked unless the community police The information included the date and system,b Abt Associates did not publicize officer is in the office. time, the type of report produced, and the existence of the system to persons c. Rocha, Joseph, “Groups Get the Latest Crime the type of data used. The logs were site not affiliated with either the CCP project Data,” The Hartford Courant, October 28, 1998. specific rather than user specific; that is, or community organizations or citizen a user identification or login name was groups. The system was the subject of two d. Mason, Johnny, “Residents Monitor Police not required to use the system because articles in the main Hartford newspaper,c,d Through Internet,” The Hartford Courant, February 16, 1999: B4.

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

and a number of drug arrests were differently about their neighborhoods ship with the police department, 6 of made as a result. when they had a comprehensive data- 17 targeted users who regularly used base of actual and suspected criminal the system indicated that the system Finally, one neighborhood organizer activity than they did when their per- helped improve the relationship; the used the NPS system to measure the ef- ception of crime was based only on re- remaining 11 indicated that the system fect of a major multiyear initiative with ports in the media and what they heard did not change the relationship. From which his organization was involved. from friends and neighbors? On this the police perspective, CSOs working For several years, his organization issue, regular NPS users expressed a in neighborhoods where the NPS sys- worked with Federal, State, and city variety of viewpoints. A majority of tem was used most frequently were officials to secure funding and garner respondents indicated that the NPS pleased to see the community have ac- political support to demolish a public maps and reports did not affect their cess to police data and appreciated the housing development in the neighbor- view of their neighborhoods: 11 of 17 work that problem-solving committees hood. One year after the demolition, regular users indicated that the level and others did to identify problem ar- the organizer used the NPS system to of reported crime displayed in the NPS eas and other crime problems. address concerns that the relocation of maps and reports was about what they public housing residents to other parts expected, and 9 of 17 said the NPS did Overall, among community organiza- of the neighborhood led to crime in- not change the extent to which they tions that regularly used the NPS sys- creases. The organizer said the NPS felt safe. Among respondents who said tem, 13 of 17 targeted users believed system enabled him to be factual the NPS system changed their view of the NPS system was “an important and precise with the organization’s their neighborhood, some indicated source of information.” The remaining supporters and city policymakers the change was positive, while others four said the system was “useful once about changes in the neighborhood. said the change was negative. Some in a while.” In general, community or- regular users saw less crime in the ganizations appear to subscribe to the Overall effects maps and reports than they expected, notion that “information is power.” As primarily because of how the media the leader of one community organiza- How did targeted NPS users and con- tion put it, “Our effectiveness depends sumers react to the NPS output? Did portrayed the crime situation. How- ever, as one community leader put on our knowledge.” Armed with infor- the maps and reports tell them any- mation, community organizations will thing they did not already know? it, “Some people are in denial about crime—they don’t want to think about be more effective in educating neigh- Although neighborhood activists were [it],” and, consequently, the maps and borhood residents, interacting with the already aware of the most problematic reports may have made them feel that police and other city agencies, and properties and locations in their neigh- the situation was worse than they pre- carrying out other specific tasks and borhoods, results of the survey of regu- viously believed. projects. In particular, several veteran lar NPS users indicated they learned community activists commented that much about their areas from the sys- Improving police-community interac- “having the data” ensures that com- tem. For example, 12 of 17 regular tion was an important objective for munity groups “will have a seat at the NPS users and consumers gained a both the Hartford Police Department table,” meaning they can make a posi- much better understanding of crime and many of the community organiza- tive contribution to problem-solving conditions in their neighborhoods, and tions involved in the project. All efforts and that their voice should not 13 of the 17 gained a much better parties hoped that providing the com- be ignored by the city. understanding of where the crime hot munity with the same data that the po- spots were in their neighborhoods. lice had would create a common frame Policy implications of reference for problem identification Of related interest was whether resi- and solving. In the end, the project Community policing requires the ac- dents’ perception of neighborhood can perhaps claim partial success in tive participation of the community, safety and quality of life were affected improving police-community relations. both in terms of formal and informal after they viewed the NPS maps and When asked whether the NPS system partnerships with the police and in reports; that is, did residents feel changed their organization’s relation- initiating its own crime prevention

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

strategies. One effective way to in- However, it is important to note that maps and reports. One option was to develop crease the community role in commu- what made the NPS system useful for a custom mapping application using a commer- cial mapping software package, such as MapInfo nity policing is routine two-way community-based problem solving or ArcView. However, to avoid having to pay for information sharing. The NIJ-funded in Hartford was the extensive infra- commercial mapping packages for all the com- effort described in this Brief documents structure that the city had in place munity organizations, the research team decided to develop a stand-alone executable program one approach by one police department to support the system, including with mapping capabilities provided by a royalty- to sharing computerized police informa- neighborhood-based problem-solving free third-party ActiveX software component. tion with the community. More police committees, veteran community orga- 4. The Federal Bureau of Investigation’s departments across the country are nizers, and a supportive environment Uniform Crime Reports list Part I crimes as making crime statistics and maps avail- in the city government and police de- homicide, rape, robbery, aggravated assault, able, primarily via police department partment. Without such an infrastruc- larceny, burglary, and auto theft. Web sites, which offer a variety of ture, it is probable that Hartford’s NPS 5. The data collection period ranged from 4 to data and features. Some offer static system not only would have been used 9 months, depending on when the NPS system maps or statistics, while others allow less but would have been used for very was installed at the site. users to formulate queries; some pro- different purposes. Specifically, in- vide aggregate data, while others stead of serving as a tool for organizing Thomas Rich is an associate and provide incident-level data. block watches, supporting neighbor- project director at Abt Associates hood organizations, and supporting Inc. This research was supported The Hartford experience highlights the neighborhood-based problem solving, different ways community groups have by the National Institute of Justice the system would have been used by and the Office of Community Ori- analyzed various data. Although all individuals interested in knowing regular users of the NPS system made ented Policing Services under grant crime conditions in their area but less number 95–IJ–CX–0079 and coop- extensive use of crime information, likely to share information from NPS some placed the highest value on erative agreement number 97–IJ– with others or act on that information CX–K017 to Abt Associates. arrest information and others closely to effect neighborhood change. This examined citizen calls-for-service is not to say that self-education about information. The map display was gen- crime conditions is not worthwhile, Findings and conclusions of the research erally the most popular report format reported here are those of the author and do but that the potential for neighborhood not necessarily reflect the official position or but still accounted for less than half of improvement is much greater when policies of the U.S. Department of Justice. the report runs, emphasizing that maps the primary system users are con- are only one way to look at police data. nected to organizations that operate The National Institute of Justice is a Appropriately sorted lists, graphs, and in an environment conducive to component of the Office of Justice time period comparisons are also es- neighborhood-based problem solving. Programs, which also includes the Bureau sential. In short, the Hartford experi- of Justice Assistance, the Bureau of Justice ence suggests that police departments Statistics, the Office of Juvenile Justice and Notes hoping to use their Web sites to pro- Delinquency Prevention, and the Office for Victims of Crime. vide data to the public in support of 1. Maltz, M.D., A. Gordon, and W. Friedman, Mapping Crime in Its Community Setting: Event their community policing initiatives Geography Analysis, New York: Springer- This and other NIJ publications can be should also include extensive query Verlag, 1991. found at and downloaded from the NIJ capabilities and multiple report for- Web site (http://www.ojp.usdoj.gov/nij). 2. One approach to developing a multilayer mats. New Web development tools, information support system for community Additional information about mapping particularly map-based tools, allow for policing is described in Block, Carolyn crime may be found at the Crime Map- this versatility. In addition, data from Rebecca, “The Geo-Archive: An Information ping Research Center Web site (http:// agencies other than police depart- Foundation for Community Policing,” Crime Mapping and Crime Prevention, ed. David www.ojp.usdoj.gov/cmrc). ments (e.g., corrections agencies, Weisburd and Tom McEwen, Monsey, property data, public health agencies, New York: Criminal Justice Press, 1998. NCJ 185333 schools) could be included to enable 3. At the beginning of the project, a variety of users to examine possible links be- options were considered for the software that tween these datasets and crime data. community organizations would use to produce

11