Using ESRI’s Districting Tool in Policing

From the Geography and Public Safety Bulletin



Environmental Systems Research Institute’s
(ESRI) districting tool1 is a free extension
for ESRI’s ArcGIS Desktop that allows analysts
to create new police districts in a city or region.
The redistricting tool initially was designed to
help governments draw congressional district
lines or create school districts, but has been
co-opted to help police departments outline
new beats. It helps them draw boundaries
and review the outcomes (e.g., predict new
workloads and crime rates) for the new districts.
The tool provides summary statistics for every
boundary selection. It does not automatically
create districts from preexisting data; rather, it
can be used to aid police communication and
make geographic information systems (GIS) a
central part of the redistricting process.
Preparing the Data
The ESRI districting tool requires an agency
to have GIS data entered as a polygon feature
class2 with attribute data. These polygons must
be smaller than the intended districts. Analysts
should use polygons that coincide with naturally
occurring boundaries and main thoroughfares.
Police departments must divide a city or region
into discrete boundaries to balance service and
distribute responsibilities into organizational
branches. This task requires departments to join
years of incident data (the point feature class)
to the chosen polygon feature class. This helps
ensure that a proportional number of incidents
is allotted to each geographic region.
The process of joining incident data to
polygons in ArcGIS to achieve a count of
incidents in each region is the same process that
an analyst would use to create a choropleth3
map that depicts different incident counts. To
join incident data to a polygon feature class, an
analyst should do the following:
1. Right-click the polygon layer...
(More Here)

UMD PhD Research: Broken Windows Policing: A Randomized Experimental Evaluation of its Impacts on Disorder, Fear and Crime in Three Cities

The RPD is providing Crime data and committing to this GIS-centric Criminology study with the University of Maryland. This is Josh Hinkle's PhD project.



David Weisburd, Josh Hinkle, Christine Famega, and Justin Ready
Over the last two decades, "Broken Windows Policing" has become a central component of police strategies to combat crime and disorder. But surprisingly, Broken Windows Policing itself has not been subject to sustained empirical examination. In this project, we address this knowledge gap by conducting a randomized, experimental evaluation of Broken Windows Policing in three cities in the San Bernardino Valley area of California. Questions addressed in this study will be whether broken windows policing reduced fear, made residents feel safer, and increased collective community efficacy. This project is being funded by the National Institute of Justice and is conducted in conjunction with researchers at the University of Maryland and California State University, San Bernardino.

Over the last two decades, "Broken Windows Policing" has become a central component of
police strategies to combat crime and disorder. But surprisingly, Broken Windows Policing itself
has not been subject to sustained empirical examination. We propose to develop a block
randomized experimental evaluation of Broken Windows Policing in three cities in the San
Bernardino Valley area of California. Specifically, we would test the impact of a Broken
Windows Policing approach upon approximately 180 street segments (divided equally into
treatment and control locations) that evidence relatively high levels of disorder at the outset of
the study. Our study would examine the impacts of Broken Windows Policing on the actual
and perceived level of minor offenses and disorder at targeted locations; its effects on fear of
crime and perceptions of police legitimacy in the areas that are targeted; and possible impacts on
crime during the study period. The experimental treatment would be delivered over a 6-month
period by special units formed within each of the three police departments. At the outset of the
study, site visits will be made by the Broken Windows Policing Unit of each department to the
"treatment" segments in their jurisdiction to assess the levels and types of disorder problems
present. After this assessment, the unit will meet and develop treatments to deliver to each
segment. The treatments available will include issuance of citations and arrests for disorderly
behaviors and minor crimes; repairs to sidewalks, streetlights and other state maintained items in the treatment segments; issuance of citations for code violations in an effort to induce repairs and clean ups; and general clean-up efforts in the treatment street segments. Our study would rely on two main types of data collection: official data on disorder and crime drawn from the police, and measures of fear of crime, perceptions of disorder and crime, and related features of citizen attitudes toward police legitimacy and community drawn from a panel survey of residents in the study area. Official data on disorder and crime would be provided by the police departments participating in the study. The panel survey will be conducted via telephone at two waves by selecting 10 households from each street segment in the study prior to the intervention, and then re-interviewing the same respondents after the intervention period.

Thanks Josh!

Flash for Animated Crime Maps

This is a earlier example (much earlier- 2002) of using animation to tell a story within a mapping environment. This is a gallery of animated maps showing burglary, assaults, all incidents, and vehicle thefts year-by-year. Also included is an example of an hour-by-hour analysis of traffic collisions.

Methodology
This is completed by exporting individual layers into Macromedia (Now Adobe) Flash. Flash is a perfect medium for this because it is able to handle raster and vectors with different levels of compression. In these (old...) examples, the densities(rasters) are vectorized, reducing the size substantially. What would be a 30-100MB avi file exported from ArcScene is an 80kb swf file.
I teach a class about this, and have since (2005) developed a way to incorporate Modelbuilder to create and symbolize the densities automatically.

Geography & Public Safety: Using Geographic Information Systems to Support Broken Windows Policing

When New York City police began
targeting graffiti crews and turnstile
jumpers instead of the city’s rampant violence
and crack use, many critics objected. But when
their strategy gave way to success, police began
seriously considering broken windows policing
as an excellent crime-reduction policy. Today,
police can use geographic information system
tools to reduce blight and graffiti in a city and
enhance broken windows policing strategies.
Broken windows policing is a theory
developed by George Kelling and James
Wilson. It suggests that if minor crimes and
disorders are targeted, major crime will be
prevented. In other words, an environmental
context sets norms of acceptable behavior—if
one window is broken, vandals will think that
breaking windows is an acceptable behavior.
Journalist Malcolm Gladwell highlights
these ideas in his book, The Tipping Point,
illustrating that many relatively benign factors
can combine to produce a “tip” in forces at
play. Criminal behavior or public disorder
grows exponentially rather than following
a straight linear path; thus, targeting small
crimes and blight can have an outstanding
effect on city crime prevention.

Crime analysts can respond to blight and
graffiti issues within a city in many different
ways. Maps using different data sources help
analysts consider the environmental factors at
play and make crime prevention a major factor
in a city’s planning and development. When
creating a plan for citywide crime prevention,
geographic information system (GIS)
analysts must combine officers’ observations
of neighborhood crime and disorder with
external data and policing initiatives.
Riverside’s Graffiti Abatement Partnership

(This article refers to Riverside CA's Graffiti abatement program. It further highlights a server-based system that utilizes GPS-enabled cameras to take images and data in the field. Public works and the PD work on the same business process. It's a great highlight of city departments working together for problem-solving initiatives).
More here...

Geography & Public Safety: Using ModelBuilder for Geographic Information System Tasks

This was published in the National Institute of Justice's Geography and Public Safety Bulletin:


Crime analysis requires a Geographic Information System (GIS)—a software program for capturing, managing, and analyzing geographic and spatial data—to process data and create data products (e.g., printed maps, pamphlets) for a wide audience. Crime data can change daily, which makes producing maps and other products a daunting task for many analysts. Automating regular tasks will save time and allow for more sophisticated tools for analysis. Crime analysts need a mapping program that allows them to regularly update data in a timely fashion. Environmental Systems Research Institute’s (ESRI) ModelBuilder is a component of a type of ArcMap that allows for drag-and-drop development of complex geoprocessing tasks (e.g., placing points on the map based on addresses, summarizing data). It helps crime analysts expedite their daily geoprocessing tasks for map production.

Advancing Earlier Modeling Tools

ESRI’s ModelBuilder is a drag-and-drop process modeling environment that comes standard with ArcMap version 9.x. ModelBuilder is an advancement of ArcView 3.x’s Geoprocessing Wizard Extension. Where the Geoprocessing Wizard was limited to tasks like spatially joining, merging, and clipping data, ESRI introduces all of ArcMap’s functional capabilities (e.g., spatial analysis, geoprocessing) into the Toolbox. A large number of different tools and tasks can be strung together in the literal sense of a systems technology: data go in, a process occurs, and data come out. The potential to string complex tasks together and automate them to run regularly makes ModelBuilder an attractive option for any Crime Analysis Unit.

The rest of the article, and a series of other articles can be found at
http://www.ojp.usdoj.gov/nij/maps/gpsBulletin-2008v1.pdf.pdf